What changed between 17 Sept 2026 and 18 Sept 2026
Two dates, one scope: new and retired models, price and context changes, new benchmark leaders, new papers, provider and hardware changes. Built from the change log keyed on when things occurred — nothing is inferred, and the URL is the report.
Scope: the whole atlas
1,006 events in the window · 1,316 claims superseded · price rows opened 21 / closed 19 · entities at 17 Sept 2026: 9,277 → at 18 Sept 2026: 9,972. The entity lists are capped at 200 by the API — per-type counts below are “of the first 200”. Events are keyed on occurred_at (effective date when known) and exclude back-filled history unless include_backfill=1; new_entities uses first_seen_at. new_benchmark_leaders compares the primary-group leader computed from results observed by each date.
New models
New models 2
| Model | Organization | Params | Context | Openness | Released | First seen |
|---|---|---|---|---|---|---|
| Grok Voice Transcribe 1.0 | xAI | — | — | — | — | 18 Sept 2026 |
| Grok Voice Transcribe 2.0 | xAI | — | — | — | — | 18 Sept 2026 |
Retired models
Retired models 0
No model was retired between these dates
Price changes
Price changes 9
| Date | Model | Provider | Input / 1M | Output / 1M | Source |
|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | OpenRouter | $0.943→$1.6+69.6% | $1.89→$3.2+69.6% | openrouter.ai | |
| Z.ai GLM 5.2 | OpenRouter | $1.4→$0.487−65.2% | $4.4→$1.56−64.5% | openrouter.ai | |
| gpt-oss-120b | OpenRouter | $0.037→$0.15+305.4% | $0.17→$0.60+252.9% | openrouter.ai | |
| DeepSeek V4 Pro 0423 | OpenRouter | $1.6→$0.943−41.0% | $3.2→$1.89−41.0% | openrouter.ai | |
| DeepSeek V4 Flash 0423 | OpenRouter | $0.07→$0.089+26.6% | $0.14→$0.177+26.6% | openrouter.ai | |
| Kimi K3 | OpenRouter | $3→$2.1−30.0% | $15→$10.95−27.0% | openrouter.ai | |
| Kimi K3 | OpenRouter | $2.1→$2−4.8% | $10.95→$11.2+2.3% | openrouter.ai | |
| Mistral Large 3 2512 | OpenRouter | $0.50→$0.55+10.0% | $1.5→$1.65+10.0% | openrouter.ai | |
| DeepSeek V4 Flash Vision Exp | OpenRouter | $0.22→$0.216−2.0% | $0.66→$0.647−2.0% | openrouter.ai |
Context changes
Context changes 3
| Date | Model | Organization | Context window | Source | History |
|---|---|---|---|---|---|
| Z.ai GLM 5.2 | Z.ai (Zhipu AI) | 131.1K tokens→163.8K tokens+25.0% | openrouter.ai | claims → | |
| DeepSeek V4 Flash Vision Exp | DeepSeek | 943.7K tokens→262.1K tokens−72.2% | openrouter.ai | claims → | |
| gpt-oss-120b | OpenAI | 118K tokens→65.5K tokens−44.4% | openrouter.ai | claims → |
New benchmark leaders
New benchmark leaders 0
Benchmarks whose primary-group leader (computed from results observed by each date) differs between 17 Sept 2026 and 18 Sept 2026.
No benchmark changed leader between these dates
New papers
New papers 197
Provider changes
Provider changes 2
| Date | Model | Provider | Input / 1M | Output / 1M | Source |
|---|---|---|---|---|---|
| DeepSeek V4 Flash (0731) | OpenRouter | —→$0 | —→$0 | openrouter.ai | |
| Pareto | OpenRouter | —→$2.5 | —→$7.5 | openrouter.ai |
Hardware changes
Hardware changes 0
No hardware change between these dates
Other property changes
Other property changes 200
Openness, licence, status, parameters, release-date and other material property changes (context changes are listed above).
phoenix: latest version changed from arize-phoenix: v20.13.0 to arize-phoenix: v20.14.0
Latest versionarize-phoenix: v20.13.0→arize-phoenix: v20.14.0githubmilvus: latest version changed from milvus-2.6.24 to client/v3.0.0
Latest versionmilvus-2.6.24→client/v3.0.0githubpydantic-ai: latest version changed from v2.44.0 (2026-09-16) to v2.45.0 (2026-09-17)
Latest versionv2.44.0 (2026-09-16)→v2.45.0 (2026-09-17)githubclaude-code: latest version changed from 2.1.275 to 2.1.276
Latest version2.1.275→2.1.276githubcodex: latest version changed from 0.156.0-alpha.1 to 0.156.0-alpha.2
Latest version0.156.0-alpha.1→0.156.0-alpha.2githublitellm: latest version changed from 1.102.0-rc.2 to 1.103.0-dev.2
Latest version1.102.0-rc.2→1.103.0-dev.2githubgemini-cli: latest version changed from Release v0.62.0-nightly.20260917.g6a466a7e2 to Release v0.62.0-nightly.20260918.g9450ade79
Latest versionRelease v0.62.0-nightly.20260917.g6a466a7e2→Release v0.62.0-nightly.20260918.g9450ade79githubpytorch: latest version changed from trunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] … to trunk/779a26b7bd8b34ba3c90091b67315bda7ff7c1cc
Latest versiontrunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] Port CPython 3.13 test_fstring to Dynamo te→trunk/779a26b7bd8b34ba3c90091b67315bda7ff7c1ccgithubletta: latest version changed from 0.16.8 to 0.32.13
Latest version0.16.8→0.32.13pypiopenai-agents: latest version changed from 0.22.2 to 0.22.3
Latest version0.22.2→0.22.3pypilancedb: latest version changed from LanceDB v0.40.0-beta.2 to LanceDB v0.39.0
Latest versionLanceDB v0.40.0-beta.2→LanceDB v0.39.0githubclaude-code: latest version changed from 2.1.274 to 2.1.275
Latest version2.1.274→2.1.275githubcodex: latest version changed from 0.155.0-alpha.16 to 0.156.0-alpha.1
Latest version0.155.0-alpha.16→0.156.0-alpha.1githublangchain: latest version changed from langchain-typesafe==0.0.1a1 to langchain-typesafe==0.0.1a2
Latest versionlangchain-typesafe==0.0.1a1→langchain-typesafe==0.0.1a2githubopenai-python: latest version changed from 3.14.1 to 3.15.0
Latest version3.14.1→3.15.0githubjax: latest version changed from JAX v0.11.1 to JAX v0.11.2
Latest versionJAX v0.11.1→JAX v0.11.2githubpytorch: latest version changed from viable/strict/1789667453 to trunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] …
Latest versionviable/strict/1789667453→trunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] Port CPython 3.13 test_fstring to Dynamo tegithubZ.ai GLM 5.2: max output tokens changed from 131072 to 163840
Max output131.1K tokens→163.8K tokensopenroutergpt-oss-120b: max output tokens changed from 117964 to 65536
Max output118K tokens→65.5K tokensopenrouterDeepSeek V4 Flash Vision Exp: max output tokens changed from 943718 to 262144
Max output943.7K tokens→262.1K tokensopenrouterphoenix: latest release at changed from 2026-09-16T22:42:58+00:00 to 2026-09-18T02:52:36+00:00
Latest release16 Sept 2026→18 Sept 2026githubphoenix: latest release tag changed from arize-phoenix: v20.13.0 to arize-phoenix: v20.14.0
Latest release tagarize-phoenix: v20.13.0→arize-phoenix: v20.14.0githubmilvus: latest release tag changed from milvus-2.6.24 to client/v3.0.0
Latest release tagmilvus-2.6.24→client/v3.0.0githubmilvus: latest release at changed from 2026-09-16T10:51:14+00:00 to 2026-09-18T04:31:53+00:00
Latest release16 Sept 2026→18 Sept 2026githubpydantic-ai: latest release at changed from 2026-09-17T04:03:52+00:00 to 2026-09-18T04:31:12+00:00
Latest release17 Sept 2026→18 Sept 2026githubpydantic-ai: latest release tag changed from v2.44.0 (2026-09-16) to v2.45.0 (2026-09-17)
Latest release tagv2.44.0 (2026-09-16)→v2.45.0 (2026-09-17)githubclaude-code: latest release tag changed from v2.1.275 to v2.1.276
Latest release tagv2.1.275→v2.1.276githubclaude-code: latest release at changed from 2026-09-17T22:33:17+00:00 to 2026-09-18T02:12:26+00:00
Latest release17 Sept 2026→18 Sept 2026githubcodex: latest release tag changed from 0.156.0-alpha.1 to 0.156.0-alpha.2
Latest release tag0.156.0-alpha.1→0.156.0-alpha.2githubcodex: latest release at changed from 2026-09-18T00:31:15+00:00 to 2026-09-18T05:04:39+00:00
Latest release18 Sept 2026→18 Sept 2026githublitellm: latest release at changed from 2026-09-16T20:29:51+00:00 to 2026-09-18T02:42:04+00:00
Latest release16 Sept 2026→18 Sept 2026githublitellm: latest release tag changed from v1.102.0-rc.2 to v1.103.0-dev.2
Latest release tagv1.102.0-rc.2→v1.103.0-dev.2githubgemini-cli: latest release at changed from 2026-09-17T01:25:44+00:00 to 2026-09-18T01:23:37+00:00
Latest release17 Sept 2026→18 Sept 2026githubgemini-cli: latest release tag changed from Release v0.62.0-nightly.20260917.g6a466a7e2 to Release v0.62.0-nightly.20260918.g9450ade79
Latest release tagRelease v0.62.0-nightly.20260917.g6a466a7e2→Release v0.62.0-nightly.20260918.g9450ade79githubTensorRT-LLM: latest release at changed from 2026-09-14T08:17:08+00:00 to 2026-09-18T03:14:53+00:00
Latest release14 Sept 2026→18 Sept 2026githubpytorch: latest release at changed from 2026-09-18T00:46:12+00:00 to 2026-09-18T06:16:08+00:00
Latest release18 Sept 2026→18 Sept 2026githubpytorch: latest release tag changed from trunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] … to trunk/779a26b7bd8b34ba3c90091b67315bda7ff7c1cc
Latest release tagtrunk/aebbfeca12cd865da461e2adbe4b09b7ef90d8aa: [dynamo] Port CPython 3.13 test_fstring to Dynamo te→trunk/779a26b7bd8b34ba3c90091b67315bda7ff7c1ccgithubAVTrace: Diagnosing Audio-Visual Temporal Reasoning in Omni Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperCan Vision-Language Models Judge Olympic Diving? From Reasoning to Scores in Zero-Shot Action Quality Assessment
Can Vision-Language Models Judge Olympic Diving? From Reasoning to Scores in Zero-Shot Action Quality Assessment: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv LinePilot Digitizer: Line-Plot Recovery with Manual and Automatic Calibration: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivRiemannian--Lorentz Fusion of Vision Transformers and State-Space Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivA Multi-Modal Generative Model for Tomato Disease Leaves Understanding: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperScientific Image Quality Assessment via Multi-modal Retrieval-Augmented Generation
Scientific Image Quality Assessment via Multi-modal Retrieval-Augmented Generation: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Federated Learning Framework for Privacy-Preserving Kidney Stone Detection: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivRegion-Level Policy Optimization for Fine-grained MLLM Perception: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv- Property changedPaperPACE: Precise AI Cinematic Expression: A Typed Specification for Script-Grounded Previsualization and Geometric Conformance
PACE: Precise AI Cinematic Expression: A Typed Specification for Script-Grounded Previsualization and Geometric Conformance: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Astronex-World 1.0: Real-Time Interactive World Model Foundation: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperPointEvent: Rethinking Event-based Tiny Object Detection via Serialized Motion Evidence Accumulation
PointEvent: Rethinking Event-based Tiny Object Detection via Serialized Motion Evidence Accumulation: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperCross-Modal Attention Acts as a Frequency Filter: Why Verbose Prompts Improve Robustness in Vision-Language Models
Cross-Modal Attention Acts as a Frequency Filter: Why Verbose Prompts Improve Robustness in Vision-Language Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperBridging Modalities on the Cortex: Surface-based MRI to PET Translation with a Diffusion Bridge
Bridging Modalities on the Cortex: Surface-based MRI to PET Translation with a Diffusion Bridge: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperTask-Oriented Semantic Feature Transmission for Multi-Task Satellite Remote Sensing over Low-SNR Channels
Task-Oriented Semantic Feature Transmission for Multi-Task Satellite Remote Sensing over Low-SNR Channels: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv CleanVideo: Adaptive Concept Erasure for Text-to-Video Diffusion Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperTinyCNN: A 193K-Parameter Network for On-Device Plant Disease Detection, with a Cross-Dataset Robustness Diagnosis
TinyCNN: A 193K-Parameter Network for On-Device Plant Disease Detection, with a Cross-Dataset Robustness Diagnosis: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperFast Cross-Strength Multi-Contrast Brain MRI Translation using Latent Bridge Matching
Fast Cross-Strength Multi-Contrast Brain MRI Translation using Latent Bridge Matching: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperTouchSight: Bare-Handed Tactile Prediction from Egocentric Video via Generative Visual Augmentation
TouchSight: Bare-Handed Tactile Prediction from Egocentric Video via Generative Visual Augmentation: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv WeVisDoc: From Coverage to Capability for Robust End-to-End Document Parsing: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv- Property changedPaperWhen Do Language-Grounded Explanations Help? A Graph-Bottleneck for Farm Monitoring Interpretable Sheep Facial Pain
When Do Language-Grounded Explanations Help? A Graph-Bottleneck for Farm Monitoring Interpretable Sheep Facial Pain: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Cross-Architecture Foundation-Model Distillation for Edge Flood Segmentation: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis
ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Paint-Anything: Unified Any-Color Control for Image Generation and Editing: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivFAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observations: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivModality Discrepancy Transformer for Ambivalence and Hesitancy Recognition: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperVABench: Measuring Embodied Spatial Intelligence through Visual Demonstrations, Active Perception, and Metric Control
VABench: Measuring Embodied Spatial Intelligence through Visual Demonstrations, Active Perception, and Metric Control: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperG3AR: Graph-Guided Neural Visual Geometry for Scalable Multi-Sequence Aerial Registration
G3AR: Graph-Guided Neural Visual Geometry for Scalable Multi-Sequence Aerial Registration: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperG3AR: Graph-Guided Neural Visual Geometry for Scalable Multi-Sequence Aerial Registration
G3AR: Graph-Guided Neural Visual Geometry for Scalable Multi-Sequence Aerial Registration: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv An Empirical Study of Harness Design for Coding Agents: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivLayer-wise Curriculum Learning for Efficient LLM Compression: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperYNU-HPCC at SemEval-2025 Task 11: Bridging the Gap in Text-Based Emotion Using Multiple Prediction Headers
YNU-HPCC at SemEval-2025 Task 11: Bridging the Gap in Text-Based Emotion Using Multiple Prediction Headers: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperRandomized SVD Approximations for Spectral Co-Clustering of Word-Document Matrices
Randomized SVD Approximations for Spectral Co-Clustering of Word-Document Matrices: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Why Pretraining Fails to Share Cross-Lingual Knowledge: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxivHow to Guide Your Language Flow: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperFor Your Eyes Only: Evaluating Coordination Between Isolated Language Model Instances
For Your Eyes Only: Evaluating Coordination Between Isolated Language Model Instances: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Compressed Active Subspaces for Scalable Bayesian Inference: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperCliniCIRCA: A Modular LLM Framework for Constructing Longitudinal Mental Health Patient Journeys from Raw EHR Narratives
CliniCIRCA: A Modular LLM Framework for Constructing Longitudinal Mental Health Patient Journeys from Raw EHR Narratives: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Form Over Content In Gradient-Based Data Attribution Methods: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperSemantic Layer Induction from Raw Telemetry via Hierarchical LLM and RAG Abstraction
Semantic Layer Induction from Raw Telemetry via Hierarchical LLM and RAG Abstraction: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Evolution or Illusion? Rethinking Evaluation in LLM Evolutionary Search: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperZarya: A Hybrid Autoregressive--Masked Diffusion Language Model with Flexible Training and Dual-Mode Inference
Zarya: A Hybrid Autoregressive--Masked Diffusion Language Model with Flexible Training and Dual-Mode Inference: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv PetriBench: Benchmarking LLM Reasoning over Dynamic State Spaces: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperKoNeoBench: A Curated Evaluation Dataset for LLM Understanding of Korean Neologisms
KoNeoBench: A Curated Evaluation Dataset for LLM Understanding of Korean Neologisms: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperPast, Future, All at Once: Mitigating Stability-Plasticity Dilemma via Post-hoc JANUS Rectification
Past, Future, All at Once: Mitigating Stability-Plasticity Dilemma via Post-hoc JANUS Rectification: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperEPIG-Tree: Compute-Optimal Branching for Gradient-Efficient Reinforcement Learning
EPIG-Tree: Compute-Optimal Branching for Gradient-Efficient Reinforcement Learning: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Dynamic Generalized Gromov-Wasserstein Optimal Transport: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperCorrect Now, Insufficient Later: Auditing Update Sufficiency in Context Compression
Correct Now, Insufficient Later: Auditing Update Sufficiency in Context Compression: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperReading Emotions in the Token Space: Discriminative Adaptation of SpeechLLMs for Emotion Recognition
Reading Emotions in the Token Space: Discriminative Adaptation of SpeechLLMs for Emotion Recognition: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Local Sparsity Enables Unsupervised LLM Safety Detection: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperQUALS: Corpus Equilibrium for Universal Forecasting via Pattern Quantization and Learnability Synchronization
QUALS: Corpus Equilibrium for Universal Forecasting via Pattern Quantization and Learnability Synchronization: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperSoftTri: Smooth Triangular Membership Functions for Adaptive Fuzzy Inference Systems
SoftTri: Smooth Triangular Membership Functions for Adaptive Fuzzy Inference Systems: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Scene-Conditioned Relation Routing for urban cellular activity forecasting: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperIs It Still Worth Training a Classical Model in the Era of LLMs? A Crossover Benchmark on Tabular Data
Is It Still Worth Training a Classical Model in the Era of LLMs? A Crossover Benchmark on Tabular Data: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperLens: Bringing the Right Semantic Perspective into Focus for Training-Free Multimodal Representation Learning
Lens: Bringing the Right Semantic Perspective into Focus for Training-Free Multimodal Representation Learning: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperLabeled Incidence Structures for Native Transformer Modeling of Text, Knowledge Graphs, and Hypergraphs
Labeled Incidence Structures for Native Transformer Modeling of Text, Knowledge Graphs, and Hypergraphs: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Xeno-Interpretability: Investigating the Alien Minds of LLMs: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxivStress-testing Alignment Midtraining: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperSCGFM-ART: Amortized Relational Transport for Structure-Centric Graph Foundation Models
SCGFM-ART: Amortized Relational Transport for Structure-Centric Graph Foundation Models: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperDeep Learning-Based Classification of Cognitive and Resting States Using Electroencephalography Signals
Deep Learning-Based Classification of Cognitive and Resting States Using Electroencephalography Signals: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv Edustories: A Collection of Real-world Case Studies from Classroom Practices: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxivAutoData: Agentic Search for Pre-training Data Selection: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivMitigating Retaliatory Algorithmic Collusion in Repeated Games: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxivWhat Do We Expect from LLMs? Mapping the Design of LLM Benchmarks: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperPosition: It is Time to Virtualize Foundation Models with a Self-evolving Operating System Layer
Position: It is Time to Virtualize Foundation Models with a Self-evolving Operating System Layer: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperCompositional Reasoning in Language Models under Reinforcement Learning Post-Training
Compositional Reasoning in Language Models under Reinforcement Learning Post-Training: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperSafety Beyond the Interface: Detecting Harm via Latent States in Large Language Models
Safety Beyond the Interface: Detecting Harm via Latent States in Large Language Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperQVAC Genesis III: A Large-Scale, High-Quality Open Synthetic STEM Corpus for Efficient Language Model Pre-Training
QVAC Genesis III: A Large-Scale, High-Quality Open Synthetic STEM Corpus for Efficient Language Model Pre-Training: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv LLM-as-an-Improver: Turning Verification into Better Candidates: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperAn Architecture for Long-Horizon Agents: Levels, Ticks and Cascaded Intelligence
An Architecture for Long-Horizon Agents: Levels, Ticks and Cascaded Intelligence: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperEconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data
EconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Self Improvement via Fast Tree-search: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperWhen Hiring Becomes Agent-Mediated: Evaluating Access and Recurrence in Two-Agent R\'esum\'e Screening
When Hiring Becomes Agent-Mediated: Evaluating Access and Recurrence in Two-Agent R\'esum\'e Screening: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperFrom Intent to Action: Benchmarking LLM Safety in Vehicle Voice Command Authorization
From Intent to Action: Benchmarking LLM Safety in Vehicle Voice Command Authorization: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperWhen2Think: Learning Difficulty-Aware Length Control for Efficient Hybrid Reasoning Models
When2Think: Learning Difficulty-Aware Length Control for Efficient Hybrid Reasoning Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Chronicle: Cut-Point Replay for Regression Testing of LLM Agents: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperHarm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations
Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperRetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv AutoResearch: Insight In, Hallucination Out: authors changed from ["Haoyang Zhang", "Jiahao Li", "Junjie Wang", "Qumeng Sun… to ["Bruno", "Haoyang Zhang", "Jiahao Li", "Qumeng Sun", "Xi…
AuthorsHaoyang Zhang, Jiahao Li, Junjie Wang, Qumeng Sun, Xiang Liu, Xiao Zhang, Yiming Ren→Bruno, Haoyang Zhang, Jiahao Li, Qumeng Sun, Xiang Liu, Xiao ZhangarxivAutoResearch: Insight In, Hallucination Out: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxiv- Property changedPaperGenerating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data
Generating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data: published at changed from 2026-09-12T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published12 Sept 2026→18 Sept 2026arxiv - Property changedPaperGenerating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data
Generating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv Atria Dawn: The Dawn of Agentic Superintelligence: authors changed from ["Aierpanjiang Simayi", "Alexander Lam", "Bing Shao", "Bi… to ["Aierpanjiang Simayi", "Alexander Lam", "Bing Shao", "Bi…
AuthorsAierpanjiang Simayi, Alexander Lam, Bing Shao, Binze Hu, Bo Li, Bowen Lai, Chengfeng Zhao, Chengjun Pan, Chenya Gu, Ding Han, Feiyang Hao, Guanting Dong, Guoqiang Wang, Haijun Lv, Hanchen Wang, Hao Xiang, Haodong Chen, Haoran Zhao, Haotan He, Honglin Guo, Houcheng Min, Huayu Sha, Huicheng Jiang, Huijie Lv, Jiabei Chen, Jiahang Lin, Jiahao Chen, Jiahao Kuang, Jiajie Jin, Jianmin Qian, Jiantao Qiu, Jianwen Luo, Jiaqiang Li, Jiaxiang Liu, Jiayang Shi, Jiayi Chen, Jiazhen Jiang, Jie Hou, Jifei Shan, Junhao Huang, Junlin Shang, Kun Cai, Lizhi Cui, Ningsheng Ma, Peng Li, Pengbo Wang, Qilong Li, Qiming Ge, Qisheng Su, Qiye Cai, Qiyuan Peng, Rui Wang, Ruojun Zhou, Shihan Dou, Shuainan Liu, Shufan Jiang, Shuhao Xing, Sirui Song, Tao Gui, Tong Liu, Wanghan Xu, Wanying Ren, Wenhui Tian, Wenzhe Tan, Xiaoxuan Peng, Xiaoyu Xing, Xin Deng, Xinchen Xie, Xinmeng Che, Xinyu Lu, Xinyu Yang, Xu Guo, Yajie Yang, Yang Shu, Yanjiang Liu, Yanyi Luo, Yi Dong, Yi Liu, Yicheng Chen, Yicheng Zou, Yiding Wang, Yining Li, Youheng Dai, Yuhui Wang, Yujia Chen, Yunhua Zhou, Yuxiao Song, Yuyang Hu, Zerun Ma, Zhe Sun, Zhecan Wang, Zhichao Sun, Zhiheng Xi, Zhongbo Tian, Zhonghang Lu, Zhuohui Sheng, Zhuoqun Li, Zijian Hu, Zixian Huang, Zixin Shang→Aierpanjiang Simayi, Alexander Lam, Bing Shao, Binze Hu, Bo Li, Bowen Lai, Chengfeng Zhao, Chengjun Pan, Chenya Gu, Ding Han, Dongsheng Zhu, Feiyang Hao, Guanting Dong, Guoqiang Wang, Haijun Lv, Hanchen Wang, Hao Xiang, Haodong Chen, Haoran Zhao, Haotan He, Honglin Guo, Houcheng Min, Huayu Sha, Huicheng Jiang, Huijie Lv, Jiabei Chen, Jiahang Lin, Jiahao Chen, Jiahao Kuang, Jiajie Jin, Jianmin Qian, Jiantao Qiu, Jianwen Luo, Jiaqiang Li, Jiaxiang Liu, Jiayang Shi, Jiayi Chen, Jiazhen Jiang, Jie Hou, Jifei Shan, Junhao Huang, Junlin Shang, Kun Cai, Lizhi Cui, Ningsheng Ma, Peng Li, Pengbo Wang, Penghao Zhao, Qilong Li, Qiming Ge, Qisheng Su, Qiye Cai, Qiyuan Peng, Rui Wang, Ruojun Zhou, Shihan Dou, Shuainan Liu, Shufan Jiang, Shuhao Xing, Sirui Song, Tao Gui, Tong Liu, Wanghan Xu, Wanying Ren, Wenhui Tian, Wenzhe Tan, Xiaoxuan Peng, Xiaoyu Xing, Xin Deng, Xinchen Xie, Xinmeng Che, Xinyu Lu, Xinyu Yang, Xu Guo, Yajie Yang, Yang Shu, Yanjiang Liu, Yanyi Luo, Yi Dong, Yi Liu, Yicheng Chen, Yicheng Zou, Yiding Wang, Yining Li, Youheng Dai, Yujia Chen, Yunhua Zhou, Yuxiao Song, Yuyang Hu, Zerun Ma, Zhe Sun, Zhecan Wang, Zhichao Sun, Zhongbo Tian, Zhonghang Lu, Zhuohui Sheng, Zhuoqun Li, Zijian Hu, Zixian Huang, Zixin ShangarxivAtria Dawn: The Dawn of Agentic Superintelligence: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxivAtria Dawn: The Dawn of Agentic Superintelligence: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv- Property changedPaperAquiLLM: Evaluating Faithfulness in Open-Weight RAG-LLM Systems for Scientific Research
AquiLLM: Evaluating Faithfulness in Open-Weight RAG-LLM Systems for Scientific Research: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxiv - Property changedPaperAquiLLM: Evaluating Faithfulness in Open-Weight RAG-LLM Systems for Scientific Research
AquiLLM: Evaluating Faithfulness in Open-Weight RAG-LLM Systems for Scientific Research: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv Can We Do Interpretable NLI with Graphs Based on Atomic Propositions?: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv- Property changedPaperThe Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction
The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperThe Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction
The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv - Property changedPaperCompositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows
Compositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows: authors changed from ["Ashwini Kurady", "Rajesh Gupta", "Sri Sai Charith Grand… to ["Ashwini Kurady", "Rajesh Gupta", "Sri Sai Charith Grand…
AuthorsAshwini Kurady, Rajesh Gupta, Sri Sai Charith Grandhi, Sumit Mamoria→Ashwini Kurady, Rajesh Gupta, Sri Sai Charith Grandhi, Sumit Kumararxiv - Property changedPaperCompositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows
Compositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperCompositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows
Compositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv - Property changedPaperDomain Elastic Transform: Bayesian Function Registration for High-Dimensional Scientific Data
Domain Elastic Transform: Bayesian Function Registration for High-Dimensional Scientific Data: published at changed from 2026-09-12T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published12 Sept 2026→18 Sept 2026arxiv - Property changedPaperTime-Aware Diffusion based on Preference Disentanglement for Generative Recommendation
Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperExplicit Language Memory for Long-Horizon Planning in Vision-Language-Action Models
Explicit Language Memory for Long-Horizon Planning in Vision-Language-Action Models: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperNo-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers
No-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers: authors changed from ["Adam Doupe", "Aditya Maheshbhai Gabani", "Chang Zhu", "… to ["Adam Doup\\'e", "Aditya Maheshbhai Gabani", "Chang Zhu"…
AuthorsAdam Doupe, Aditya Maheshbhai Gabani, Chang Zhu, Divij Handa, Hongkai Chen, Hulin Wang, Jie Hu, Pratham Hegde, Ruoyu Wang, Siyu Liu, Souradip Nath, Tiffany Bao, Yan Shoshitaishvili, Yibo Liu, Zehua Zhang, Zhuoer Lyu→Adam Doup\'e, Aditya Maheshbhai Gabani, Chang Zhu, Divij Handa, Hongkai Chen, Hulin Wang, Jie Hu, Pratham Hegde, Ruoyu Wang, Siyu Liu, Souradip Nath, Tiffany Bao, Yan Shoshitaishvili, Yibo Liu, Zehua Zhang, Zhuoer Lyuarxiv - Property changedPaperNo-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers
No-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers: published at changed from 2026-09-12T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published12 Sept 2026→18 Sept 2026arxiv - Property changedPaperNo-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers
No-Box Vulnerability Analysis: Description-only Detection of Indirect Prompt Injection Vulnerabilities in MCP Servers: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv A Mathematical Theory of Pragmatic Information: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv- Property changedPaperUFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation
UFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation: published at changed from 2026-09-14T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published14 Sept 2026→18 Sept 2026arxiv - Property changedPaperUFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation
UFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv Legislating World-Model-Based Planning with Legal Reasoning: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxivLegislating World-Model-Based Planning with Legal Reasoning: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv- Property changedPaperIWC-Bench: Evaluating Web Application Generation from a Software Testing Perspective
IWC-Bench: Evaluating Web Application Generation from a Software Testing Perspective: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxiv - Property changedPaperIWC-Bench: Evaluating Web Application Generation from a Software Testing Perspective
IWC-Bench: Evaluating Web Application Generation from a Software Testing Perspective: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv - Property changedPaperCSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models
CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models: authors changed from ["Chong Ma", "Jian Zhu", "Jianjun Zhang", "Taiyi Su", "Ti… to ["Chong Ma", "Jian Zhu", "Jianjun Zhang", "Taiyi Su", "Ti…
AuthorsChong Ma, Jian Zhu, Jianjun Zhang, Taiyi Su, Tianbin Liu, Yi Xu, Zitai Huang→Chong Ma, Jian Zhu, Jianjun Zhang, Taiyi Su, Tianbin Liu, Weiyi Lu, Yi Xu, Zitai Huangarxiv - Property changedPaperCSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models
CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperCSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models
CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv - Property changedPaperDreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation
Dreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperDreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation
Dreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv TorchCraft: Unified binder design by inverting an all-atom structure predictor: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperSteering Equilibrium Selection in Regularized Self-Play via the Reference Policy
Steering Equilibrium Selection in Regularized Self-Play via the Reference Policy: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperReproducibility is not construct validity: LLM measurement of institutionally situated communication
Reproducibility is not construct validity: LLM measurement of institutionally situated communication: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperFrom "Who Is This User?" to "What Does This Purchase Mean?": A Deployed Pipeline for Semantic User Profiling at Bank Scale
From "Who Is This User?" to "What Does This Purchase Mean?": A Deployed Pipeline for Semantic User Profiling at Bank Scale: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Geopolitical Divisions Across Languages in Large Language Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperMarginal utility, matrix factorization, and the Key-Value (KV) cache: a unified information-economic framework for sovereign geo-mining inference
Marginal utility, matrix factorization, and the Key-Value (KV) cache: a unified information-economic framework for sovereign geo-mining inference: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperPaGNet: A Panel-Aware GBDT--Neural Network for Multi-Target Corporate Tax Avoidance Proxy Forecasting
PaGNet: A Panel-Aware GBDT--Neural Network for Multi-Target Corporate Tax Avoidance Proxy Forecasting: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperSequential Contextual Fit Predicts Human Behavioural and Neural Dynamics Across Domains
Sequential Contextual Fit Predicts Human Behavioural and Neural Dynamics Across Domains: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperAI-Driven Real-Time Relay Optimisation in Smart Urban NR-V2X Networks via Learning-to-Optimise Graph Neural Networks
AI-Driven Real-Time Relay Optimisation in Smart Urban NR-V2X Networks via Learning-to-Optimise Graph Neural Networks: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv- Property changedPaperLanguage-model groups overstate consensus when replaying human deliberation on a reasoning task
Language-model groups overstate consensus when replaying human deliberation on a reasoning task: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperPAA: The Probabilistic Allen Algebra: A Generative and Complete Probabilistic Extension of Allen's Interval Relations
PAA: The Probabilistic Allen Algebra: A Generative and Complete Probabilistic Extension of Allen's Interval Relations: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperFRAUDSkill: Structured Frozen-Weight Skill Optimization for Audio Anti-Fraud Detection
FRAUDSkill: Structured Frozen-Weight Skill Optimization for Audio Anti-Fraud Detection: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperSubliminal Prompting Beyond Static Geometry: Causal Depth and Multi-Token Confounds
Subliminal Prompting Beyond Static Geometry: Causal Depth and Multi-Token Confounds: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperSampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations
Sampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperStop Removing Stopwords: How an Inherited Preprocessing Default Distorts Legal Text-as-Data
Stop Removing Stopwords: How an Inherited Preprocessing Default Distorts Legal Text-as-Data: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperVisKG-LM: Compiling Knowledge Graphs into Visual Memory for Multiple-Choice Question Answering
VisKG-LM: Compiling Knowledge Graphs into Visual Memory for Multiple-Choice Question Answering: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv - Property changedPaperAdvantage Scale Calibration Imbalance in Group-Relative Optimization under Low-Variance Rewards: Diagnosis and Bounded Recovery
Advantage Scale Calibration Imbalance in Group-Relative Optimization under Low-Variance Rewards: Diagnosis and Bounded Recovery: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Why Pretraining Fails to Share Cross-Lingual Knowledge: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperCliniCIRCA: A Modular LLM Framework for Constructing Longitudinal Mental Health Patient Journeys from Raw EHR Narratives
CliniCIRCA: A Modular LLM Framework for Constructing Longitudinal Mental Health Patient Journeys from Raw EHR Narratives: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv Chain-of-Thought Entropy as a Reliability Signal: A Preregistered Reproduction: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivEvolution or Illusion? Rethinking Evaluation in LLM Evolutionary Search: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperIntrinsic Sequence-Likelihood Confidence in Retrieval-Dominated Extractive QA: Two Pre-Specified Negatives, and What They Do and Do Not Attribute
Intrinsic Sequence-Likelihood Confidence in Retrieval-Dominated Extractive QA: Two Pre-Specified Negatives, and What They Do and Do Not Attribute: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression: authors changed from ["Anyi Xu", "B. Li", "Bangcai Lin", "Bing Xue", "BingChen… to [":", "Anyi Xu", "B. Li", "Bangcai Lin", "Bing Xue", "Bin…
AuthorsAnyi Xu, B. Li, Bangcai Lin, Bing Xue, BingCheng Xian, Bingzheng Xu, Bochao Wu, Bowei Zhang, Boyi Deng, C. C. Yu, Chao Jin, Chaofan Lin, Chen Dong, Chenbing Wang, Chenfan Feng, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chengyuan Zhang, Chenhao Xu, Chenqi Zhao, Chenze Shao, Chuhao Wang, Chuqi Zhang, Damai Dai, DeepSeek-AI, Dejian Yang, Deli Chen, Di Huang, Di Wu, Donghao Li, Erhang Li, Eric Fu, F. Zhou, Fangwei Zhou, Fangyun Lin, Fangzhou Yuan, Feiyu Xia, Fucong Dai, Guangbo Hao, Guanglin Li, Guanting Chen, Guoai Cao, Guofan Fan, Guolai Meng, Guowei Li, Haichuan Zhang, Haiyang Ma, Haiyang Shen, Han Li→:, Anyi Xu, B. Li, Bangcai Lin, Bing Xue, BingCheng Xian, Bingzheng Xu, Bochao Wu, Bowei Zhang, Boyi Deng, C. C. Yu, Chao Jin, Chaofan Lin, Chen Dong, Chenbing Wang, Chenfan Feng, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chengyuan Zhang, Chenhao Xu, Chenqi Zhao, Chenze Shao, Chuhao Wang, Chuqi Zhang, Damai Dai, DeepSeek-AI, Dejian Yang, Deli Chen, Di Huang, Di Wu, Donghao Li, Erhang Li, Eric Fu, F. Zhou, Fangwei Zhou, Fangyun Lin, Fangzhou Yuan, Feiyu Xia, Fucong Dai, Guangbo Hao, Guanglin Li, Guanting Chen, Guoai Cao, Guofan Fan, Guolai Meng, Guowei Li, Haichuan Zhang, Haiyang Ma, Haiyang Shen, Han Li, Han Yu, Han Zhang, Hangyuan Deng, Hanwei Xu, Hanxiang Xu, Hanxun Zhong, Hao Guo, Hao Jiang, Hao Li, Hao Qin, Haodong Wen, Haofen Liang, Haofeng Huang, Haohua Liu, Haoling Zhang, Haoming Luo, Haoran Yang, Haotian Xu, Haotian Yuan, Haoting Huang, Haowen Luo, Haoyang Cai, Haoyu Chen, Haozhe Ji, Hengran Zhang, Hengrui Wang, Hengxu Wu, Honghui Ding, Hongxuan Tang, Huadong Wang, Huanqi Cao, Huazuo Gao, Hui Qu, Hui Zeng, J. H. Jin, J. H. Zhang, J. X. Zou, J. Yang, Jia Yu, Jiahui Zhou, Jiajun Chen, Jialiang Huang, Jialin Zhao, Jiamin Tang, Jian Zhou, Jianan Tong, Jianwen Li, Jiaqi Zhu, Jiarui WangarxivDeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxivRelational Attention for Data-Efficient Language Modeling: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivWhat Does Privileged Information Add to On-Policy Self-Distillation?: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxivdQwen3.5: Hybrid-Attention Diffusion Language Models: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxiv- Property changedPaperRetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv JEPA-Anything: Learning Predictive Models across Different Worlds: published at changed from 2026-09-17T00:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxivEmbedding Models Measure in Peculiar Ways: arxiv announce type changed from cross to new
Arxiv announce typecross→newarxivLearn Your Own Thoughts: Abstract Token Curriculum: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxivUni-LaDiR: Latent Diffusion Unifies Multimodal Reasoning: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv- Property changedPaperMATCH: Model-Aware Tool Learning with Curriculum Scheduling and Hierarchically Gated Rewards
MATCH: Model-Aware Tool Learning with Curriculum Scheduling and Hierarchically Gated Rewards: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv - Property changedPaperDon't Mask the Environment: Observation Supervision Changes How Agents Explore Under RL
Don't Mask the Environment: Observation Supervision Changes How Agents Explore Under RL: arxiv announce type changed from new to cross
Arxiv announce typenew→crossarxiv LongWoF-Bench: Evaluating EvoMap Genes for Verifiable Long-Workflow Tasks: authors changed from ["Haoyang Zhang", "Jiahao Li", "Junjie Wang", "Qumeng Sun… to ["Bruno", "Haoyang Zhang", "Jiahao Li", "Qumeng Sun", "Xi…
AuthorsHaoyang Zhang, Jiahao Li, Junjie Wang, Qumeng Sun, Xiang Liu, Xiao Zhang, Yiming Ren→Bruno, Haoyang Zhang, Jiahao Li, Qumeng Sun, Xiang Liu, Xiao ZhangarxivLongWoF-Bench: Evaluating EvoMap Genes for Verifiable Long-Workflow Tasks: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxivCausal Analysis and Mitigation of Spurious Onsets in Full-Duplex Speech LLMs: published at changed from 2026-09-15T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published15 Sept 2026→18 Sept 2026arxivCausal Analysis and Mitigation of Spurious Onsets in Full-Duplex Speech LLMs: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxivMUSE: A Theory-Harnessed Story Engine for Vibe Narrativizing: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxivSlopShape: Identifying AI-Generated Commercial Web Content: published at changed from 2026-09-15T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published15 Sept 2026→18 Sept 2026arxivSlopShape: Identifying AI-Generated Commercial Web Content: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv- Property changedPaperVerifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering
Verifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering: published at changed from 2026-09-15T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published15 Sept 2026→18 Sept 2026arxiv - Property changedPaperVerifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering
Verifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv How Humans and LLMs Read Gender into Gender-Neutral Physical Descriptions: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxivHow Humans and LLMs Read Gender into Gender-Neutral Physical Descriptions: arxiv announce type changed from cross to replace
Arxiv announce typecross→replacearxiv- Property changedPaperRiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation
RiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxiv - Property changedPaperRiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation
RiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv - Property changedPaperRollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM Agents
Rollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM Agents: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxiv - Property changedPaperRollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM Agents
Rollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM Agents: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv SEA-LION-v4.8: A Technical Report: published at changed from 2026-09-17T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published17 Sept 2026→18 Sept 2026arxivSEA-LION-v4.8: A Technical Report: arxiv announce type changed from new to replace
Arxiv announce typenew→replacearxiv- Property changedPaperFrom Procedural Skills to Strategy Genes: Towards Experience-Driven Test-Time Evolution
From Procedural Skills to Strategy Genes: Towards Experience-Driven Test-Time Evolution: published at changed from 2026-09-16T04:00:00+00:00 to 2026-09-18T04:00:00+00:00
Published16 Sept 2026→18 Sept 2026arxiv
Other new entities
Other new entities 1
| Entity | Type | Organization | First seen |
|---|---|---|---|
| Tongyi Qianwen License | License | — | 18 Sept 2026 |
Gone entities
Gone entities 0
No entity disappeared between these dates
“Gone” means an entity present at the first date is no longer current at the second (merged or retired) — its record and history are kept. Events are keyed on occurred_at; back-filled history is excluded (tick “include backfill” to add it). Per-entity history: any entity's History tab. Atlas as of 17 Sept 2026 · Methodology →