{"total":104,"items":[{"citing_arxiv_id":"2606.31247","ref_index":259,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"FlexiSLM: A Dynamic and Controllable Frame Rate Spoken Language Model","primary_cat":"cs.SD","submitted_at":"2026-06-30T07:24:10+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"FlexiSLM is the first spoken language model supporting dynamic and controllable frame rates on speech input and output, outperforming fixed-rate 7B models at high quality and enabling faster inference at lower rates like 6.25 Hz.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.30814","ref_index":109,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"When Calibration Rankings Reverse: Accuracy-Controlled Evaluation for Fair Comparison of LLMs","primary_cat":"cs.CL","submitted_at":"2026-06-29T18:37:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Global calibration metrics like ECE are confounded by accuracy; the proposed ACE framework with three accuracy-controlled views shows many prior calibration advantages weaken or reverse.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.29844","ref_index":29,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"MATCH: Modulating Attention via In-Context Retrieval for Long-Context Transformers","primary_cat":"cs.CL","submitted_at":"2026-06-29T06:33:37+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"MATCH augments sparsified attention with an efficient in-context retrieval system to boost performance on long-range recall tasks in transformers.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.28615","ref_index":7,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"What LLMs explain is not what they believe: Evaluating explanation sufficiency under models' own input beliefs","primary_cat":"cs.LG","submitted_at":"2026-06-26T21:14:56+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Proposes SCSuff metric for evaluating LLM explanation sufficiency via model-generated alternative inputs, showing explanations are typically insufficient and predictable from hidden states.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.28044","ref_index":284,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"A Tree-of-Thoughts Inspired Hybrid Approach for Legal Case Judgement Summarization using LLMs","primary_cat":"cs.CL","submitted_at":"2026-06-26T12:46:27+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"A tree-of-thoughts inspired hybrid extractive-abstractive LLM prompt yields better legal case judgment summaries than standard extractive or abstractive prompts.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.27786","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"SHIFT: Gate-Modulated Activation Steering for Knowledge Conflict Mitigation in Retrieval-Augmented Generation","primary_cat":"cs.CL","submitted_at":"2026-06-26T07:17:11+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"SHIFT reformulates neuron editing as learnable gate modulation on under 0.01% parameters to let LLMs adaptively balance contextual and parametric knowledge during RAG generation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.27679","ref_index":31,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"From Signals to Transfer: A Factorised Study of Probe-Based Uncertainty Estimation in Large Language Models","primary_cat":"cs.CL","submitted_at":"2026-06-26T03:23:15+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A factorized study finds raw hidden states and attention features hard to beat in-domain for LLM uncertainty probes, but structured compressed features are more robust under distribution shift, with pretrained probes transferring to open-ended generation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.27361","ref_index":41,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Autoregressive Boltzmann Generators","primary_cat":"cs.LG","submitted_at":"2026-06-25T17:58:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"ArBG replaces flow-based methods with autoregressive models for Boltzmann sampling, showing gains on peptide benchmarks and a 132M-parameter model Robin cutting zero-shot energy error by over 60% on 8-residue systems.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.23591","ref_index":66,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Quantifying the Agreement Between Data-Influence and Data-Similarity to Understand LLM Behavior","primary_cat":"cs.LG","submitted_at":"2026-06-22T17:00:04+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Data-similarity and data-influence produce significantly overlapping rankings of training documents for LLM outputs, with asymmetry allowing a favorable cost-accuracy trade-off.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.22645","ref_index":12,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"All Relations Lead to Rome: Automated Knowledge Graph Creation and Question Generation","primary_cat":"cs.IR","submitted_at":"2026-06-21T19:09:54+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"ARLtR is a framework for jointly constructing knowledge graphs, embeddings, and grounded QA pairs from text, demonstrated on a Roman Empire dataset with over 19,000 entities and 8,400 QA pairs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.22511","ref_index":44,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Breaking the Likelihood Trap: Variance-Calibrated Modulation for Large Language Model Decoding","primary_cat":"cs.CL","submitted_at":"2026-06-21T14:04:51+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"VCM is a training-free decoding intervention that applies PMI-driven token elevation and variance-adaptive penalization to reduce repetitive degeneration in LLM open-ended generation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.19667","ref_index":22,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"CacheWeaver: Cache-Aware Evidence Ordering for Efficient Grounded RAG Inference","primary_cat":"cs.CL","submitted_at":"2026-06-18T00:38:46+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"CacheWeaver is a lightweight scheduling layer that orders evidence to exploit prefix caching, reducing median TTFT by 20-33% across vLLM setups while preserving answer quality.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.13392","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"MiniMax Sparse Attention","primary_cat":"cs.AI","submitted_at":"2026-06-11T14:23:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MiniMax Sparse Attention is a GQA-based block-sparse attention mechanism that selects top-k blocks independently per group and delivers 28.4x per-token compute reduction at 1M context with on-par performance plus 14.2x prefill and 7.6x decode speedups via co-designed GPU kernel.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.12397","ref_index":13,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Redesign Mixture-of-Experts Routers with Manifold Power Iteration","primary_cat":"cs.LG","submitted_at":"2026-06-10T17:57:36+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Manifold Power Iteration aligns MoE router rows with principal singular directions of experts via a power-then-retract process, with theory showing convergence and experiments on 1B-11B models showing gains.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.12117","ref_index":21,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Soft-Prompt Tuning for Fair and Efficient LLM Benchmark Evaluation","primary_cat":"cs.CL","submitted_at":"2026-06-10T14:12:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Soft-prompt tuning with 10 vectors improves format compliance on LLM benchmarks and provides a low-cost proxy for comparing base models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.12479","ref_index":14,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"ReCal: Reward Calibration for RL-based LLM Routing","primary_cat":"cs.LG","submitted_at":"2026-06-10T06:59:36+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"ReCal introduces hierarchical reward decomposition and distribution-aware optimization to address ambiguous credit assignment and optimization bias in RL-based LLM routing.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.11470","ref_index":109,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes","primary_cat":"cs.CL","submitted_at":"2026-06-09T21:59:37+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A literature survey that introduces a taxonomy for LLM reasoning paradigms, analyzes methodological trends, and synthesizes failure modes from over 300 papers.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.08804","ref_index":66,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Q-Delta: Beyond Key-Value Associative State Evolution","primary_cat":"cs.AI","submitted_at":"2026-06-07T19:49:01+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Q-Delta extends linear attention by introducing a query-conditioned delta rule that incorporates mixed key-query errors into recurrent state updates for improved stability and performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.08702","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"ConMem: Structured Memory-Guided Adaptation in Training-Free Multi-Agent Systems","primary_cat":"cs.AI","submitted_at":"2026-06-07T15:59:15+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"ConMem distills agent trajectories into structured memory cards organized in a relation-aware graph to enable training-free, relation-coordinated adaptation in LLM-based multi-agent systems.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.06087","ref_index":49,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents","primary_cat":"cs.CL","submitted_at":"2026-06-04T12:26:09+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LatentSkill uses a hypernetwork to generate LoRA adapters from textual skills, enabling weight-space storage that cuts prefill tokens and boosts agent success rates on ALFWorld and Search-QA.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.05734","ref_index":1,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"When AI Says It Feels","primary_cat":"cs.AI","submitted_at":"2026-06-04T05:49:34+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"LLMs trained via rubric-based self-rewarding RL with GRPO enhanced feeling expression and sycophancy robustness but degraded truthful QA performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.05054","ref_index":143,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Boosting Self-Consistency with Ranking","primary_cat":"cs.CL","submitted_at":"2026-06-03T16:12:30+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"RISC reformulates self-consistency answer selection as a ranking task solved by a lightweight LambdaRank model with five hand-designed features, yielding better accuracy-efficiency trade-offs than majority voting on QA benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.04302","ref_index":76,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"LazyAttention: Efficient Retrieval-Augmented Generation with Deferred Positional Encoding","primary_cat":"cs.CL","submitted_at":"2026-06-03T00:12:22+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LazyAttention kernelizes deferred positional encoding to enable zero-copy, position-agnostic KV cache reuse, delivering 1.37× lower TTFT and 1.40× higher throughput than Block-Attention under skewed document distributions while preserving output quality.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.03846","ref_index":11,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Clustered Self-Assessment: A Simple yet Effective Method for Uncertainty Quantification in Large Language Models","primary_cat":"cs.CL","submitted_at":"2026-06-02T16:25:54+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Clustered Self-Assessment groups sampled LLM responses into semantic clusters, presents clusters as multiple-choice options, and uses the LLM's assigned probabilities to those options as direct uncertainty estimates, outperforming entropy baselines with as few as two extra samples.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.03197","ref_index":5,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"MemTrain: Self-Supervised Context Memory Training","primary_cat":"cs.CL","submitted_at":"2026-06-02T05:56:24+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"MemTrain introduces two coupled self-supervised proxy tasks on Wikipedia corpora to train general context-memory capabilities in LLMs, reporting gains of up to 17.67 points on long-text and search-based QA benchmarks over direct post-training.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.02544","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"SimSD: Simple Speculative Decoding in Diffusion Language Models","primary_cat":"cs.CL","submitted_at":"2026-06-01T17:46:46+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"SimSD adds a masking strategy to enable speculative decoding in diffusion LLMs, delivering up to 7.46x throughput gains on SDAR models while preserving generation quality.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.02245","ref_index":20,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"When Knowledge Is Not Free: Cost-Aware Evidence Selection in Retrieval-Augmented Generation","primary_cat":"cs.CL","submitted_at":"2026-06-01T13:39:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Defines cost-aware RAG with evidence cost tiers and shows static selectors are brittle while agentic LLM-based selection is promising but model-dependent.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.02093","ref_index":9,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"The Role of Ambiguity in Error Prediction via Uncertainty Quantification","primary_cat":"cs.CL","submitted_at":"2026-06-01T11:20:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Disentangling input ambiguity from uncertainty quantification improves error prediction for LLMs on QA tasks, yielding over 10 PRR point gains across models and datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.01923","ref_index":56,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Resonant Context Anchoring: Decoupling Attention Routing and Signal Gain at Inference Time","primary_cat":"cs.CL","submitted_at":"2026-06-01T08:57:51+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"RCA is a training-free module that boosts input context signal strength in the residual stream of LLMs by orthogonal decoupling of attention routing from value magnitude.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.01033","ref_index":18,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"TriLens: Per-Layer Logit-Lens Entropy for White-Box Hallucination Detection","primary_cat":"cs.AI","submitted_at":"2026-05-31T05:48:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"TriLens detects hallucinations via per-layer entropy trajectories of logit-lens readouts from three internal modules across LLMs and QA benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.00881","ref_index":11,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Chunking Methods on Retrieval-Augmented Generation - Effectiveness Evaluation Against Computational Cost and Limitations","primary_cat":"cs.CL","submitted_at":"2026-05-30T20:32:08+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Empirical study claiming to be the first broad comparison of chunking methods in RAG, highlighting effectiveness, cost, and generalization limitations across scenarios.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.00683","ref_index":43,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"OCC-RAG: Optimal Cognitive Core for Faithful Question Answering","primary_cat":"cs.CL","submitted_at":"2026-05-30T11:42:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"OCC-RAG develops task-specialized SLMs (0.6B and 1.7B) via a new synthetic data pipeline for multi-hop reasoning and context faithfulness, claiming to match or exceed 2-6x larger general models on HotpotQA, MuSiQue, TAT-QA, ConFiQA, and MuSiQue-Un.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.00357","ref_index":20,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"From \"Weak\" Signals to Strong Models: Preference Delta Aggregation with LoRA Merging","primary_cat":"cs.AI","submitted_at":"2026-05-29T21:00:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Aggregating preference deltas from several weak-weaker model pairs via LoRA adapters and geometric alignment merging improves strong-model performance on reasoning and search benchmarks beyond any single delta.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.07597","ref_index":14,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Repetition Mismatch: Why Data Mixture Experiments Don't Scale and How to Fix Them","primary_cat":"cs.LG","submitted_at":"2026-05-29T06:08:57+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Repetition rate mismatch between small-scale proxies and target budgets is the main reason data mixture experiments do not scale; a subsampling procedure that equalizes repetition rates recovers optimal mixtures from 1/16-scale experiments.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29727","ref_index":35,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Bastion: Budget-Aware Speculative Decoding with Tree-structured Block Diffusion Drafting","primary_cat":"cs.LG","submitted_at":"2026-05-28T10:21:34+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"BASTION is a budget-aware speculative decoding framework with adaptive tree-structured block diffusion drafting that reports up to 6.61x speedup and 39% improvement over block-diffusion baselines.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29224","ref_index":10,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agents","primary_cat":"cs.CL","submitted_at":"2026-05-28T01:23:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Web retrieval degrades safety alignment in LLM agents, with relevance activating vulnerabilities including a Safe Source Paradox where oppositional content increases harmful compliance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.28721","ref_index":10,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?","primary_cat":"cs.AI","submitted_at":"2026-05-27T16:39:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LiveBrowseComp shows search agents rely on intrinsic knowledge on standard benchmarks, with scores dropping 25-40 points and closed-book accuracy below 2% on questions about facts from the prior 90 days.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.27996","ref_index":35,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Reward Bias Substitution: Single-Axis Bias Mitigations Redirect Optimization Pressure","primary_cat":"cs.AI","submitted_at":"2026-05-27T05:40:22+00:00","verdict":"ACCEPT","verdict_confidence":"MODERATE","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Single-axis reward bias mitigations redirect optimization pressure to correlated proxies, and audit-distribution scoring produces identical observables for successful mitigation, bias substitution, and overcorrection.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.27164","ref_index":22,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Query Symbolically or Retrieve Semantically? A Dataset and Method for Semi-Structured Question Answering","primary_cat":"cs.AI","submitted_at":"2026-05-26T15:22:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"MODERATE","novelty_score":6.0,"formal_verification":"none","one_line_summary":"DualGraph combines semantic textual KGs with symbolic KGs for semi-structured QA and introduces the SpecsQA benchmark, outperforming baselines on both open and specification questions.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.26366","ref_index":23,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Automatic Layer Selection for Hallucination Detection","primary_cat":"cs.AI","submitted_at":"2026-05-25T22:28:23+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"FEPoID automatically selects optimal or near-optimal intermediate layers for hallucination detection across LLM architectures and tasks, outperforming prior criteria and baselines, with an added truncation step that further improves performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.27445","ref_index":17,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"RAGe: A Retrieval-Augmented Generation Evaluation Framework","primary_cat":"cs.IR","submitted_at":"2026-05-23T17:46:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"RAGe is a modular evaluation framework that correlates retrieval and generation quality with hardware constraints to recommend optimal RAG components for specific datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.21413","ref_index":17,"ref_count":2,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Teaching AI Through Benchmark Construction: QuestBench as a Course-Based Practice for Accountable Knowledge Work","primary_cat":"cs.AI","submitted_at":"2026-05-20T17:09:56+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"QuestBench is a student-constructed benchmark of 256 questions on which current deep research AI systems achieve a mean pass rate of 16.85% and a best-case rate of 57.58%.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20490","ref_index":45,"ref_count":3,"confidence":0.88,"is_internal_anchor":false,"paper_title":"ECUAS$_n$: A family of metrics for principled evaluation of uncertainty-augmented systems","primary_cat":"cs.AI","submitted_at":"2026-05-19T20:55:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"ECUAS_n is a parameterized family of proper scoring rules for jointly assessing prediction accuracy and uncertainty quality in automated decision systems.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.17989","ref_index":4,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Predictive Prefetching for Retrieval-Augmented Generation","primary_cat":"cs.CL","submitted_at":"2026-05-18T07:45:27+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Introduces predictive prefetching for RAG that anticipates retrieval needs several tokens ahead via three components, reporting up to 43.5% latency reduction and 62.4% TTFT improvement while preserving answer quality.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14754","ref_index":4,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"XDomainBench: Diagnosing Reasoning Collapse in High-Dimensional Scientific Knowledge Composition","primary_cat":"cs.AI","submitted_at":"2026-05-14T12:19:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"XDomainBench shows LLMs suffer systematic reasoning collapse as domain composition order increases due to direct difficulty and interaction-amplified failures.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14449","ref_index":13,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"When Answers Stray from Questions: Hallucination Detection via Question-Answer Orthogonal Decomposition","primary_cat":"cs.LG","submitted_at":"2026-05-14T06:44:18+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"QAOD projects away question-aligned directions from answer representations to isolate domain-agnostic factuality signals, enabling efficient hallucination detection with top in-domain AUROC and up to 21% better OOD transfer.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.12227","ref_index":2,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"A Recipe for Long-Context Reasoning in Large Language Models via On-Policy Optimization and Distillation","primary_cat":"cs.CL","submitted_at":"2026-05-12T15:04:18+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Combines GRPO with teacher-guided on-policy distillation and introduces LongBlocks dataset to yield more stable long-context reasoning than either method alone.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"data-efficient approach to bootstrap long-context behavior and can be viewed as a form of imitation learning (Ross et al., 2011) for language models. Formally, given an expert dataset D of input-output pairs (x, y), we optimize the model parameters θ by minimizing the token-level negative log-likelihood: LSFT(θ) =E (x,y)∼D \" − |y| ∑ t=1 logπ θ(yt |x,y