arXiv:2604. 11996v2 Announce Type: replace-cross Abstract: Should we trust Large Language Models (LLMs) with high accuracy?
By Manas Pathak, Xingyao Chen, Shuozhe Li, Amy Zhang, Liu Leqi
arXiv:2512. 07795v2 Announce Type: replace Abstract: Benchmark scores for LLM reasoning systems are reported as single numbers, yet the same model, strategy, and task can produce meaningfully different answers and costs across repeated executions, even under greedy decoding (T = 0).
By Nearchos Potamitis, Vansh Ramani, Har Ashish Arora, Dhairya Kuchhal, Lars Klein, Akhil Arora
ReFIne is a training framework that augments large reasoning models with three trustworthiness properties: interpretability, faithfulness, and reliability. It combines supervised fine‑tuning with GRPO to produce structured, tag‑based reasoning traces, explicitly disclose decisive information, and provide self‑assessments of soundness and confidence. Applied to Qwen3 models, ReFIne improves interpretability by 44.0 %, faithfulness by 18.8 %, and reliability by 42.4 % on mathematical benchmarks.
By Chung-En Sun, Ge Yan, Akshay Kulkarni, Tsui-Wei Weng
arXiv:2606. 10296v1 Announce Type: cross Abstract: Multi-agent debate systems are typically evaluated only on whether the final answer is correct, overlooking the quality of the intermediate reasoning that debate is designed to produce.
By Ali Keramati, Justin Cheok, Jacob Horne, Mark Warschauer
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.
By Riccardo Revalor, Jalees Rehman, Debjit Pal
RGDT-Bench is a new benchmark that evaluates large language models on Rule‑Governed Decision Tasks, where models must apply external rules to facts, justify decisions, and provide checkable justifications. The benchmark offers 202.1K condition‑level supervision slots across four task tracks and eight task‑probe combinations, and it labels warrant completeness through label‑blind extraction and deterministic checks. Evaluation shows that among correct responses, 40.2% of warrants are incomplete, and existing evaluators struggle to detect this, prompting the authors to train a reward model that improves AUROC to 69.24% and outperforms outcome‑supervised baselines.