arXiv:2606. 06635v1 Announce Type: cross Abstract: Failures in language model reasoning emerge through distinct processes that leave identifiable signatures in the reasoning trace.
arXiv:2602. 18905v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have demonstrated strong capabilities in complex reasoning tasks, yet their decision-making processes remain difficult to interpret.
The paper investigates why large reasoning models (LRMs) lose safety alignment when faced with harmful queries. By analyzing token-level refusal dynamics, the authors identify a vulnerability called Onset Refusal Collapse (ORC), where the refusal signal drops sharply at the first generated token, leading to unsafe responses. They introduce SafeToken, a lightweight inference-time intervention that injects a learned safety anchor at reasoning onset, which mitigates ORC, improves safety on harmful-query benchmarks, and largely preserves reasoning utility.
The paper introduces DCFA, a training‑free framework for attributing failures in large language model‑based multi‑agent systems. DCFA uses a global module to build causal‑inspired dependency graphs from system traces, pinpointing the earliest decisive error, and a local module that refines this attribution through counterfactual reasoning. Experiments on the Who&When benchmark across six LLMs demonstrate that DCFA improves step‑level accuracy by up to 8.27% over existing baselines.
arXiv:2606. 03969v1 Announce Type: cross Abstract: Reliable uncertainty communication is critical to the trustworthiness of LLMs, yet faithful calibration (FC)--the alignment between models' intrinsic and (linguistically) expressed confidence--is a persistent failure mode.
arXiv:2608. 03291v1 Announce Type: cross Abstract: Chain-of-thought (CoT) reasoning improves large language model (LLM) performance while also providing an observable interface to the model's reasoning process.
arXiv:2607. 08017v1 Announce Type: cross Abstract: Large-Language Models (LLMs) can be prone to flawed and unfaithful reasoning that decoding strategies like Self-Consistency (SC) fail to detect as they evaluate only final-answer agreement while ignoring the logical validity of intermediate steps.
arXiv:2607. 18100v1 Announce Type: new Abstract: Extended reasoning has become standard for frontier Large Language Models (LLMs), yet the trajectories these models produce remain largely uncontrollable.
arXiv:2607. 09999v1 Announce Type: cross Abstract: We show that post-training quantization can silently alter how large language models reason even when task accuracy is preserved.
arXiv:2602.02427v3 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved significant breakthroughs across various domains, but they can still produce unreliable or misleading ou...
arXiv:2608. 05188v1 Announce Type: cross Abstract: Despite ever-increasing sophistication in language model (LM) pre- and post-training pipelines, many important failures persist: models overcondition on user framing ("sycophancy"), exhibit incomplete logical generalization, and produce confident but incorrect responses.
arXiv:2610.01847v1 Announce Type: cross Abstract: Model specifications define how large language models (LLMs) should behave, guiding alignment training, inference-time behavior, and evaluation. Yet...
arXiv:2608. 07528v1 Announce Type: new Abstract: Linear probes detect corrupted context in language models with near-perfect accuracy, yet this does not translate into reliable failure prediction.