arXiv:2606. 02902v1 Announce Type: cross Abstract: Deep reinforcement learning (DRL) is increasingly applied to de novo molecular design, but choices in data, rewards, and evaluation can yield uneven performance across disease areas and chemotypes.
arXiv:2606. 30887v1 Announce Type: cross Abstract: Large language models show promise for mental health support, yet therapeutic quality improves only when evaluation functions as an actionable control signal rather than a passive metric.
The paper investigates how large language models (LLMs) used as judges in absolute scoring tasks exhibit systematic biases that compromise reliability. It shows that a judge’s task accuracy strongly predicts both its judging accuracy and its directional bias, yet more capable examinee models consistently receive more lenient judgments. To mitigate these biases, the authors propose a calibrated weighted majority voting (WMV) ensemble that estimates judges’ error rates from inter-judge agreement patterns, achieving near-oracle performance without labeled data and improving both accuracy and fairness.
The paper introduces a two‑dimensional construct validity framework for evaluating large language models (LLMs) as judges, defining invariance (S) and sensitivity (R) to construct‑preserving and construct‑changing edits. Experiments across seven judges and four domains reveal high invariance (average S = 0.945) but low sensitivity (average R = 0.319), with sensitivity varying by edit type. Audits of public label sets show that surface‑only predictors can reproduce a substantial portion of labels, underscoring that high agreement does not guarantee construct validity.
The paper investigates two strategies for improving large language model (LLM) evaluation: specialized judge weights and rule‑based deferral policies. Experiments on nearly 100,000 rubric‑conditioned samples show that correct rubrics boost accuracy, while incorrect ones hurt it, and that splitting training data into criterion‑specific experts can severely degrade performance unless the experts are warm‑started from a unified model. The authors demonstrate that lightweight deferral cascades can match or exceed the accuracy of larger standalone judges at a fraction of the compute cost, and they provide practical design rules for building efficient, reliable LLM evaluators.
arXiv:2607. 28788v1 Announce Type: new Abstract: Clinical diagnosis at hospital admission must be made rapidly from limited, incomplete evidence.
arXiv:2607. 08535v1 Announce Type: cross Abstract: An LLM-as-judge score can move even when the candidate responses stay fixed, simply because the evaluator has changed.
The study introduces Malaria-Instruct, a curated instruction-following dataset for malaria virtual screening, and evaluates five open-source large language models (Gemma-2, TxGemma, and LlaSMol-Mistral) against classical machine learning baselines and proprietary models. Fine‑tuned LLMs outperform all baselines, with TxGemma-9B achieving the highest ROC‑AUC (0.731 ± 0.005) and LlaSMol-Mistral-7B delivering the best enrichment factor (EF@1% ≈ 4.99). The results demonstrate that domain‑specific fine‑tuning and chemistry‑aware pretraining are essential for reliable discrimination, positioning fine‑tuned open‑source LLMs as a resource‑efficient alternative for antimalarial virtual screening.
arXiv:2510. 17532v2 Announce Type: replace-cross Abstract: Predicting cancer treatment outcomes requires models that are both accurate and interpretable, particularly in the presence of heterogeneous clinical data.
The paper introduces an LLM-as-a-Judge framework for evaluating the outputs of an agentic drug discovery assistant, ChatInvent, deployed at AstraZeneca. It defines four quality dimensions—Completeness, Relevancy, Structural Clarity, and Scope Adherence—alongside deterministic Tool Call Correctness checks, and validates the judge against five expert annotators. After optimizing the best-performing judge with few-shot demonstrations, alignment with human majority votes improves from 0.80 to 0.86, and the framework reveals that informal question phrasing does not degrade output quality.
Converting free-text radiology reports into structured labels supports cohort building, quality assurance, and monitoring of clinical imaging models, but the strongest label extractors are hosted prop...
arXiv:2608.29168v1 Announce Type: new Abstract: The LLM-as-a-Judge paradigm has emerged as a scalable alternative to human evaluation. However, single-model judges are limited by their inherent model...