The paper presents a calibrated evidence inference model fine‑tuned from Qwen3‑4B‑Base using low‑rank adapters and a classification head to classify trial outcomes as significant increase, significant decrease, or no significant difference. Trained on 1,431 open‑access PubMed Central articles, the released model (Apache 2.0) achieves an expected calibration error of 0.017, a log loss lower than Gemma 4 31B‑it in zero‑shot settings, and a macro‑F1 of 0.925 on the test split. Accepting its top‑80 % most confident predictions yields only a 2.6 % error rate, and ablation studies show that article context influences answers while the title alone still provides useful signal.
By Johann Emmanuel Li
The paper introduces a new taxonomy for benchmark contamination that categorizes leakage by the mitigation it defeats—direct, derivative, temporal, distributional, and acquired—covering both training‑time and evaluation‑time scenarios. It proposes a four‑field disclosure protocol to record contamination status alongside benchmark scores, and provides a JSON schema, validator, and examples. An empirical study of 41 documents using a pre‑registered instrument shows limited reporting of contamination types and variable reliability, highlighting gaps in current disclosure practices.
By Johanna Angulo, V\'ictor Yeste, Hector Espinos-Morato
The paper investigates when post‑training quantization (PTQ) calibration helps or hurts text embedders, focusing on the amount of calibration (dose) and whether to calibrate at all. Across three models and 161 calibration runs at 4‑bit weight quantization, the effect ranges from significant gains to catastrophic degradation, driven mainly by the functional displacement of the quantized model. A validation‑gated abstention rule is proposed that avoids harmful calibrations while capturing most benefits, highlighting the trade‑off between safety and opportunity.
By Hyojung Han, Ryangkyung Kang, Youngeun Kim, Seung-Hun Jeon
The study evaluates Jev 1.13, a non‑generative model that selects from predefined answer options, on four medical benchmarks: MetaMedQA, PubMedQA, DiagnosisArena‑MCQ, and the NEJM Case Challenges. Jev’s top‑1 accuracy matches GPT‑6 Sol with medium reasoning on PubMedQA but falls behind on MetaMedQA, DiagnosisArena‑MCQ, and NEJM cases. While Jev shows strong calibration on MetaMedQA and is fast and inexpensive, its performance on examination and complex diagnostic tasks is substantially lower, indicating the need for task‑specific validation before clinical deployment.
By Alfredo Madrid-Garc\'ia, Beatriz Merino-Barbancho
arXiv:2609.34024v1 Announce Type: cross
Abstract: Jev is a non-generative "System One" model that assigns probabilities to predefined answer options and cannot answer outside them. Its accuracy and c...
By Alfredo Madrid-Garc\'ia, Beatriz Merino-Barbancho
arXiv:2609.00654v1 Announce Type: new
Abstract: We describe the SciTrue team's participation in both subtasks of the NTCIR-19 SciClaimEval task~\cite{sciclaimeval}, which asks systems to verify scien...
By Qiming Bao, Ne\c{s}et \"Ozkan Tan, Siyuan Wang, Mark Gahegan
The paper presents a system for the MedReason 2026 challenge that tackles both multiple‑choice and open‑ended medical visual question answering using offline, containerized inference. Key findings include that comparing answer semantics rather than labels boosts retrieval‑only accuracy from 20.0 % to 57.5 % on a 200‑case holdout, and that varying the number of in‑prompt retrieved examples has minimal impact on final accuracy (93.5 %–94.0 %). The final system achieves 94.0 % MCQ accuracy on the development set and 93.20 % on the official pre‑evaluation, far surpassing the off‑the‑shelf baseline.
"whyItMatters":"The results demonstrate that semantic‑aware retrieval and careful adapter tuning can dramatically improve medical VQA performance, offering a practical approach for high‑accuracy, offline inference in clinical settings."
By Tristan Kirscher (ICube, Institut Strauss), Niklas C. Koser (CAU), Soren Pirk (CAU)
The study evaluates large language model (LLM) graders on two computer‑science exams, testing 171 configurations of closed‑ and open‑weights models. While the best LLM configuration achieved a mean absolute error of 1.64/35—better than the 2.61/35 error between two human graders—its performance was highly sensitive to the prompt. A short "strict grader" preamble caused most open‑weight models to exceed acceptable error thresholds or stop grading entirely, whereas fine‑tuning with a single LoRA adapter restored parity with human graders and reduced sensitivity to harsh prompts.
By Ali Habibullah, Yazan Alshoibi, Mohammad Alshiekh, Salman Khan, Naeemullah Khan
arXiv:2606. 05633v1 Announce Type: new Abstract: Retrieval-augmented QA pipelines often route retrieved passages through an LLM \emph{rewriter} before a smaller reader, lifting F1 by tens of points on multi-hop benchmarks; this gain is typically credited to improved evidence quality.
By Yuejie Li, Yueying Hua, Ke Yang, Li Zhang, Yueping He, Yueping He, Ruiqi Li, Bolin Chen, Tao Wang, Bowen Li, Chengjun Mao
arXiv:2610.29052v1 Announce Type: cross
Abstract: A typed decision maps evidence, a question and a fixed set of options to a probability for every option, computed in one forward pass, by reading a l...
By Yuchuan Weng, Chi Liu, Xin Chen, Srinivasan Manoharan
arXiv:2609.38021v1 Announce Type: cross
Abstract: We evaluate an auditable long-term memory system on LongMemEval-S. Its retrieval chain uses hybrid candidate retrieval, cross-encoder reranking, cove...
By Christopher J. Chanhnourack
The paper investigates how sequential knowledge editing can degrade a language model’s ability to discern reliable evidence from unreliable evidence without affecting overall accuracy. Using a conservatively tuned LoRA on Qwen2.5‑7B‑Instruct, the authors show that after 1,000 edits the model’s arbitration score for untouched facts drops by 36%, leading to higher error rates on its most confident decisions, while MMLU accuracy remains unchanged. The study also finds that in some model‑method combinations, sequential edits can reduce MMLU to chance levels even though edit success and locality remain perfect.
By Atul Anand