Legible Failures: Detecting and Repairing In-Context Binding Errors
Updated 1 h ago · first seen 11 Sept 2026
paper_01M294FP407CJX0TG0GKT81FPY
- Published
- 11 Sept 2026
- T1 · 1 h ago
- arXiv
- 2609.11216
- T1 · 1 h ago
- Category
- cs.LG
- T1 · 1 h ago
Abstract
A wrong answer does not show whether the model lacked the needed information or held it and failed to use it. On an entity-obligation binding task, a language model can emit an incorrect prompt-supplied binding while a linear probe can recover the correct one from its frozen hidden state. We measure how often this occurs across 16 public checkpoints, each evaluated with three seeds. We fit a probe on a training fold, select its layer on a validation fold, and report results on a disjoint test fold. On the trials each model gets wrong, probe accuracy exceeds the strict present-obligation baseline, 1/K = 0.125, by +0.196 (95% CI [+0.101, +0.296], bootstrapped over models). A query-entity counterfactual rules out token presence and recency. A score built from the sign of probe-output disagreement improves failure detection over the model's own confidence by +0.079 AUROC (95% CI [+0.036, +0.126]). Raw probe confidence gives no measurable improvement over model confidence. Steering the residual stream toward the probe-decoded binding, with no gold label, raises accuracy on all eight models tested by a mean of +0.168 (95% CI [+0.066, +0.280]). Where recent studies report that probe-detected errors are resistant to interventions, we find that in-context binding is a setting in which probes are actionable.
Authors 3
Manas Venkata Sai Ravulapalli, Samrath Singh Chadha, Abhinav M. Hari
Specification
- Official page
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
- Arxiv announce type
- new
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
- arXiv id
- 2609.11216
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
- Categories
- cs.LG
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
- Primary category
- cs.LG
Source:arXiv (Atom API + RSS)T1observed 1 h agohigh
- Published
- 11 Sept 2026
Source:arXiv (Atom API + RSS)T1observed 1 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
1 h ago
Conflicts
None
No models linked to this paper yet.
As of
Rewind the record: see this entity's attributes exactly as AI Atlas knew them on a given day.
Claim history · Abstract
Abstractabstract1
| Value | Valid from → to | Status | Source | Confidence | Extractor |
|---|---|---|---|---|---|
| A wrong answer does not show whether the model lacked the needed information or held it and failed to use it. On an entity-obligation binding task, a language model can emit an incorrect prompt-supplied binding while a linear probe can recover the correct one from its frozen hidden state. We measure how often this occurs across 16 public checkpoints, each evaluated with three seeds. We fit a probe on a training fold, select its layer on a validation fold, and report results on a disjoint test fold. On the trials each model gets wrong, probe accuracy exceeds the strict present-obligation baseline, 1/K = 0.125, by +0.196 (95% CI [+0.101, +0.296], bootstrapped over models). A query-entity counterfactual rules out token presence and recency. A score built from the sign of probe-output disagreement improves failure detection over the model's own confidence by +0.079 AUROC (95% CI [+0.036, +0.126]). Raw probe confidence gives no measurable improvement over model confidence. Steering the residual stream toward the probe-decoded binding, with no gold label, raises accuracy on all eight models tested by a mean of +0.168 (95% CI [+0.066, +0.280]). Where recent studies report that probe-detected errors are resistant to interventions, we find that in-context binding is a setting in which probes are actionable. | → current | current | arXiv (Atom API + RSS)T1 | high | deterministic |
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
New paper: Legible Failures: Detecting and Repairing In-Context Binding Errors
arxiv
| Source | Document | Type | Tier | Last observed | Snapshots |
|---|---|---|---|---|---|
| arXiv (Atom API + RSS) | rss.arxiv.org/rss/cs.LG | feed | T1· Official | 1 h ago | 1 |
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.