{"total":11,"items":[{"citing_arxiv_id":"2605.17694","ref_index":37,"ref_count":2,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Do LLM Agents Mirror Socio-Cognitive Effects in Power-Asymmetric Conversations?","primary_cat":"cs.CL","submitted_at":"2026-05-17T23:23:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"LLMs assigned high or low status personas in multi-turn dialogues exhibit socio-cognitive effects including language coordination, pronoun patterns, persuasion success, and compliance with unsafe requests.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14115","ref_index":19,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"When Evidence Conflicts: Uncertainty and Order Effects in Retrieval-Augmented Biomedical Question Answering","primary_cat":"cs.CL","submitted_at":"2026-05-13T21:02:24+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Conflicting biomedical evidence triggers order-dependent prediction flips in RAG LLMs, and a new abstention score combining confidence with conflict detection raises selective accuracy by 7-33 points in the hardest conditions.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.12809","ref_index":18,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces","primary_cat":"cs.LG","submitted_at":"2026-05-12T23:01:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"By the definition of the directional derivative, evaluating the Jacobian along the standard basis vectorej yields: JG(rtrain)e j = d dε G(rtrain +εe j) ε=0 = JVP(G, rtrain, ej).(17) Putting Eq.s (15)-(17) together, we obtain a computable form for the neuron-level influence: I r j (zr train, ztest) =−g ⊤ testH −1 θ2 r(j) train JVP(G, rtrain, ej).(18) Here, the JVP term quantifies the downstream gradient sensitivity, while the activation factor (r(j) train) enforces that only active circuits contribute. This directly connects data-level influence attribution to feature-level monosemanticity [Bricken et al., 2023]. The Computational Bottleneck.While Eq. (18) avoids storing the full Jacobian, its evaluation"},{"citing_arxiv_id":"2605.11865","ref_index":30,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Variance-aware Reward Modeling with Anchor Guidance","primary_cat":"stat.ML","submitted_at":"2026-05-12T09:46:26+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Anchor-guided variance-aware reward modeling uses two response-level anchors to resolve non-identifiability in Gaussian models of pluralistic preferences, yielding provable identification, a joint training objective, and improved RLHF performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.09716","ref_index":29,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Medical Model Synthesis Architectures: A Case Study","primary_cat":"cs.AI","submitted_at":"2026-05-10T19:30:16+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"MedMSA framework retrieves knowledge via language models then builds formal probabilistic models to produce uncertainty-weighted differential diagnoses from symptoms.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.09675","ref_index":4,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"CodeClinic: Evaluating Automation of Coding Skills for Clinical Reasoning Agents","primary_cat":"cs.AI","submitted_at":"2026-05-10T17:45:01+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"CodeClinic benchmark demonstrates that LLM-generated Python skill libraries from clinical guidelines enhance consistency and reduce token consumption by up to 40% compared to zero-shot approaches on MIMIC-IV based tasks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.08439","ref_index":13,"ref_count":2,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Can Language Models Identify Side Effects of Breast Cancer Radiation Treatments?","primary_cat":"cs.CL","submitted_at":"2026-05-08T20:02:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"LLMs exhibit sensitivity to small input changes and systematically under-recall rare and long-term side effects when listing radiation toxicities for breast cancer, with major gains from grounding in clinician-curated references.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.05893","ref_index":43,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Logic-Regularized Verifier Elicits Reasoning from LLMs","primary_cat":"cs.CL","submitted_at":"2026-05-07T09:03:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LOVER creates an unsupervised logic-regularized verifier that reaches 95% of supervised verifier performance on reasoning tasks across 10 datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.03895","ref_index":13,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"From Data Lifting to Continuous Risk Estimation: A Process-Aware Pipeline for Predictive Monitoring of Clinical Pathways","primary_cat":"cs.LG","submitted_at":"2026-05-05T15:51:43+00:00","verdict":"CONDITIONAL","verdict_confidence":"MODERATE","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A process-aware pipeline using logistic regression on prefix representations from 4,479 COVID-19 cases predicts ICU admission with AUC 0.906, improving from 0.642 early to 0.942 later in pathways.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.18976","ref_index":19,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"STAR-Teaming: A Strategy-Response Multiplex Network Approach to Automated LLM Red Teaming","primary_cat":"cs.CL","submitted_at":"2026-04-21T01:58:09+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"STAR-Teaming uses a Strategy-Response Multiplex Network inside a multi-agent framework to organize attack strategies into semantic communities, delivering higher attack success rates on LLMs at lower computational cost than prior methods.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2404.18416","ref_index":194,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Capabilities of Gemini Models in Medicine","primary_cat":"cs.AI","submitted_at":"2024-04-29T04:11:28+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Med-Gemini sets new records on 10 of 14 medical benchmarks including 91.1% on MedQA-USMLE, beats GPT-4V by 44.5% on multimodal tasks, and surpasses humans on medical text summarization.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}