Skip to content
AI Atlas
PaperActive

TRACE: Training Reasoning Agents for Causal Exploration with Synthesized Rewards

arxiv.org/abs/2609.10315

quality89

Updated 5 h ago · first seen 11 Sept 2026

paper_01M294GKH8J975T2NW8M1CDWBE

Published
11 Sept 2026
T1 · 5 h ago
arXiv
2609.10315
T1 · 5 h ago
Category
cs.AI
T1 · 5 h ago

Abstract

Reinforcement learning with verifiable rewards (RLVR) has advanced language-model reasoning in domains such as mathematics and code, where objective answers are inexpensive to check. Diagnostic reasoning over complex data lacks this advantage: establishing the true cause of an anomaly often requires costly expert investigation and may remain ambiguous after the fact. We ask whether this asymmetry of verification can instead be engineered. We sample an intervention, inject it into a controlled simulator, and generate the observations it would produce. The hidden intervention provides an oracle label and objective reward, while the agent must still investigate noisy, confounded, and distributed evidence. We instantiate this approach in TRACE, a digital-advertising diagnostic environment with 12 root causes and fine-grained segment attribution. Agents investigate each episode using Python and SQL and must identify both the root cause and, when applicable, the affected segment assignment. On a held-out 235-episode test set, the strongest prompted baseline, Claude Opus 5, reaches 0.686 FullAttr@1. Supervised fine-tuning raises Qwen3.5-35B-A3B from 0.159 to 0.637, and subsequent RL with synthesized rewards reaches 0.757, outperforming all evaluated prompted baselines, including frontier closed-source models and a prompted Qwen3.5-122B-A10B model. The resulting policy also uses substantially fewer tool calls than the prompted 35B base. These results provide evidence that access to a scalable, objective training signal can be a more important constraint than model scale alone. More broadly, simulation-based verification can make otherwise ambiguous diagnostic reasoning tasks amenable to scalable reinforcement learning.

Authors 3

Rui Sun, Zhan Shi, Bing He

Specification

Official page

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

Arxiv announce type
new

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

arXiv id
2609.10315

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

Categories
cs.AI, cs.LG

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

PDF

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

Primary category
cs.AI

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

Published
11 Sept 2026

Source:arXiv (Atom API + RSS)T1observed 5 h agohigh

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

9

Source tiers

T19

Freshest observation

5 h ago

Conflicts

None