The paper introduces SEAV, a verification‑centric framework for evaluating jailbreak attempts against large language models. SEAV decomposes responses into ordered steps and checks both validity and correctness using LLM‑as‑a‑judge and retrieval‑grounded verification. The method reduces false positives by 14.9 percentage points on a strategic‑dishonesty diagnostic and reclassifies 22.1–51.0% of previously successful jailbreaks as invalid across multiple benchmarks.
By Qilong Wu, Sahil Wadhwa, Pranab Mohanty, Giri Iyengar, Varun Chandrasekaran
arXiv:2604. 03904v2 Announce Type: replace-cross Abstract: Large language models (LLMs) often produce confident but incorrect answers, in part because standard evaluation incentives reward guessing over expressing uncertainty.
By Haotian Zong, Binze Li, Yufei Long, Sinyin Chang, Jialong Wu, Gillian K. Hadfield
PROOF is a benchmark that profiles the reliability of object-level facts in instruction-tuned language models by converting a frozen Wikidata snapshot into 18,486 English multiple-choice questions grounded in 11,779 semantic facts across 101 classes, 392 properties, and 14 domains. Each question includes an explicit "I don't know" option, a "No correct option" control, and nine controlled formulations, with 1,849 questions designed as no-correct-option traps. The study evaluates 18 open-weight model deployments on 166,374 prompts, revealing wide variability in factual accuracy, sensitivity to wording changes, and the impact of decoder perturbations.
By Andrei Chetvergov, Mikhail Solovev, Timofei Sivoraksha, Stepan Ukolov, Valeriia Kuschenko, Alexander Evseev, Sergey Bolovtsov
The paper introduces Semantic Confusion to assess how consistently large language models refuse similar prompts. It presents ParaGuard, a 10k‑prompt corpus of controlled paraphrase clusters, and proposes three token‑level metrics—Confusion Index, Confusion Rate, and Confusion Depth—to measure contradictory refusal decisions across meaning‑preserving paraphrases. Experiments show that global false rejection rates can mask local inconsistencies, revealing that refusal evaluation must consider both frequency and consistency across nearby paraphrases.
By Riad Ahmed Anonto, Md Labid Al Nahiyan, Md Tanvir Hassan
arXiv:2607. 26102v1 Announce Type: cross Abstract: Mathematical chain of thought (CoT) evaluation is commonly reduced to whether the final answer matches a reference.
By Vivek Shukla, Varun Shukla, Atul, Divya Mishra, Mehul Kumar Das
The paper investigates whether existing AI safety benchmarks, designed for large language models, are suitable for evaluating small language models (SLMs). By testing five benchmark suites on 26 open‑source SLMs with a unified scoring rubric, the authors find that ambiguous judgments dominate, especially for complex prompts and certain architectures. This ambiguity, linked to factors like lexical density and output perplexity, undermines the reliability of aggregate leaderboards and reveals a confound between model capability and perceived safety.
By Nyamtulla Shaik, Fengjun Li, Bo Luo
The paper challenges the common practice of estimating aleatoric uncertainty in large language models (LLMs) by generating multiple clarified inputs and comparing the resulting answers. It argues that answers are unnecessary, costly, and can introduce epistemic leakage, proposing instead a clarification-only method that directly assesses ambiguity from the space of plausible interpretations. Experiments on three benchmarks show the new approach improves AUROC, reduces computational cost, and yields uncertainty estimates less correlated with epistemic uncertainty.
By Omer Nahum, Niv Nayman, Jonathan Fhima, Alon Zolfi, Jeremy Levy, Shai Mazor, Paolo Favaro
arXiv:2608. 09080v1 Announce Type: cross Abstract: Large Language Models (LLMs) have achieved strong performance in medical question answering and clinical reasoning tasks.
By Maryam Tahermazandarani, Adnan Mahmood, Fahmida Islam, Quan Z. Sheng
arXiv:2607. 25600v1 Announce Type: cross Abstract: Retrieval-augmented generation improves knowledge-intensive question answering, but indiscriminate retrieval can introduce irrelevant evidence and unnecessary computation.
By Chandan Kumar Sah, Xiaoli Lian, Li Zhang
The paper introduces DEDUCE, a three‑stage framework that turns large language models into proactive error correctors by detecting input fact errors, devising correction strategies, and delivering reliable answers. It also presents MisFactQA, a dataset of factual errors, and new metrics for robustness evaluation. Experiments on TruthfulQA, FalseQA, and MisFactQA show significant gains in accuracy and error correction across Qwen, LLaMA, and Gemma models.
By Ping Wang, Xiangguo Sun, Bingbing Xu, Guocong Li, Xiaofeng Meng
The paper introduces a new evaluation framework for confidence estimation in large language models, focusing on three properties: robustness to prompt changes, stability across semantically equivalent answers, and sensitivity to semantically different answers. It demonstrates that existing confidence estimation methods perform well on robustness and stability but often fail to detect differences in answer meaning, revealing gaps in current evaluation practices. The framework aims to guide the selection of confidence estimators for practical applications.
By Yuxi Xia, Dennis Ulmer, Terra Blevins, Yihong Liu, Hinrich Sch\"utze, Benjamin Roth
arXiv:2608.22138v1 Announce Type: new
Abstract: Language models are usually judged by a single accuracy score, which does not reveal how their performance degrades as inputs are perturbed. We present...
By Samira Golsefid