{"total":174,"items":[{"citing_arxiv_id":"2606.20751","ref_index":83,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"From Sentiment to Actionable Insights: A Data-Driven Public Sentiment Analysis of Advanced Air Mobility","primary_cat":"cs.CL","submitted_at":"2026-06-18T03:07:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Applies standard sentiment classifiers and topic modeling to a large AAM discussion corpus, identifies six clusters of public concern, and lists strategies to address them.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09131","ref_index":15,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Late-Layer Fusion is Enough: Dual-Path Vision Token Routing for Multimodal Large Language Models under Visual Saturation","primary_cat":"cs.AI","submitted_at":"2026-06-08T07:28:14+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"DPVR-LF routes saturated vision tokens into a one-layer side branch after layer 4, runs text-only processing through layers 5-17, and performs late fusion at the final layer to reduce visual computation while preserving multimodal performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09881","ref_index":66,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Toward Calibrated, Fair, and accurate Deepfake Detection","primary_cat":"cs.LG","submitted_at":"2026-06-03T05:44:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.31512","ref_index":38,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Reliable Multilingual Orthopedic Decision Support from Clinical Narratives: Language-Aware Adaptation and Verification-Guided Deferral","primary_cat":"cs.CL","submitted_at":"2026-05-29T16:30:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"IndicBERT-HPA with language-aware adapters and verification-guided deferral outperforms baselines on multilingual orthopedic note classification, reaching 0.8792 Macro-F1 overall and 84.4% selective accuracy at 72.3% coverage.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.30729","ref_index":29,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"SemStruct: Contextualizing Semantic Embeddings with Structural Information for Schema Matching","primary_cat":"cs.LG","submitted_at":"2026-05-29T01:45:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"SemStruct models tables as heterogeneous graphs with GNNs on frozen PLM embeddings to incorporate row co-occurrences for schema matching and reports SOTA results on Valentine and SOTAB-SM benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.30448","ref_index":30,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Bounded Behavioral Indistinguishability for Black-Box LLM Distillation","primary_cat":"cs.LG","submitted_at":"2026-05-28T18:19:32+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Introduces (ε,q,t,A)-behavioral indistinguishability and shows via Qwen/Llama experiments that LoRA distillation boosts semantic similarity but leaves detectable behavioral differences under adversarial evaluation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29755","ref_index":17,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Rec-Distill: An Industrial Distillation Pipeline for Large-Scale Recommendation Models","primary_cat":"cs.IR","submitted_at":"2026-05-28T10:59:07+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Rec-Distill is an industrial distillation pipeline that transfers substantial performance from large-scale recommendation models to efficient students, reporting over 60% transferability and measurable business gains.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29705","ref_index":43,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"BitTP: The Lightweight Trajectory Prediction Model with BitLLM for Edge-Devices","primary_cat":"cs.AI","submitted_at":"2026-05-28T10:04:02+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"BitTP applies weight-only 1.58-bit quantization to LLM trajectory predictors, claiming improved ADE/FDE over BF16 baseline with reduced resource demands on edge devices.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29302","ref_index":10,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"ViASNet: A Video Ad Saliency Network for Predicting Dynamic Saliency and Viewer Engagement","primary_cat":"cs.CV","submitted_at":"2026-05-28T03:33:07+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"ViASNet applies a 3D U-Net architecture augmented with audio and semantic inputs to predict dynamic saliency in video ads and uses frame-wise entropy to diagnose low-engagement scenes on eye-tracked data from 151 ads.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.28767","ref_index":85,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Principled Algorithms for Optimizing Generalized Metrics in Multi-Label Learning","primary_cat":"cs.LG","submitted_at":"2026-05-27T17:23:58+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Develops H-consistent surrogate losses for generalized metrics in multi-label classification that decompose exactly in O(l) time and introduces the MMO family of algorithms.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.28483","ref_index":24,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"From Learning Resources to Competencies: LLM-Based Tagging with Evidence and Graph Constraints","primary_cat":"cs.AI","submitted_at":"2026-05-27T13:41:10+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"An LLM+BM25+graph pipeline tags learning resources to competencies with evidence spans, reaching 0.57 micro-F1 and 0.50 macro-F1 at fragment level on a 22-competency university dataset while outperforming baselines.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.27345","ref_index":2,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"MATCHA: Matching Text via Contrastive Semantic Alignment","primary_cat":"cs.CL","submitted_at":"2026-05-26T17:47:14+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"MATCHA introduces a dual-view contrastive metric measuring proximity to gold text and distance from adversarial contradictions, outperforming ROUGE and BERTScore by up to 20% on TruthfulQA and other NLP benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.23857","ref_index":12,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Strong Teacher Not Needed? On Distillation in LLM Pretraining","primary_cat":"cs.LG","submitted_at":"2026-05-22T17:16:35+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Even small or undertrained teachers improve larger LLM students via distillation with tuned loss mixing, while stronger teachers can saturate or reverse gains and distillation aids generalization more than in-domain fit.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.23482","ref_index":54,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Multimodal Distribution Matching for Vision-Language Dataset Distillation","primary_cat":"cs.CV","submitted_at":"2026-05-22T10:41:58+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MDM distills vision-language datasets via joint embedding clustering, weight-space model interpolation, and geometry-aware distribution matching on the unit hypersphere.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.23244","ref_index":114,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Convex Optimization for Alignment and Preference Learning on a Single GPU","primary_cat":"cs.LG","submitted_at":"2026-05-22T05:25:00+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"COALA applies convex optimization reformulations of neural networks to direct preference optimization, claiming single-GPU training with ~18% of DPO's TFLOPs and competitive performance on multiple datasets and models up to 8B parameters.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.22779","ref_index":22,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"FAME: Failure-Aware Mixture-of-Experts for Message-Level Log Anomaly Detection","primary_cat":"cs.SE","submitted_at":"2026-05-21T17:34:53+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"FAME achieves F1 of 98.16 on BGL and 99.95 on Thunderbird for message-level log anomaly detection using at most K=100 labels per template, reducing annotation effort by 76x while detecting anomalies from unseen EventIDs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.22738","ref_index":54,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Proxy-Based Approximation of Shapley and Banzhaf Interactions","primary_cat":"cs.LG","submitted_at":"2026-05-21T17:09:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"ProxySHAP approximates higher-order Shapley and Banzhaf interactions via tree proxies plus residual correction and a polynomial-time interventional TreeSHAP generalization for tree ensembles.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.22731","ref_index":23,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Post-Training is About States, Not Tokens: A State Distribution View of SFT, RL, and On-Policy Distillation","primary_cat":"cs.LG","submitted_at":"2026-05-21T17:03:22+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A state distribution view of post-training shows that on-policy supervision from the learner itself can outperform fixed-dataset SFT and preserve retention better than aggressive supervised updates.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.22120","ref_index":47,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Effective User-defined Keyword Spotting with Dual-stage Matching, Multi-modal Enrollment, and Continual Adaptation","primary_cat":"eess.AS","submitted_at":"2026-05-21T07:52:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"DMA-KWS achieves 97.85% AUC and 6.13% EER on LibriPhrase Hard via dual-stage CTC/QbyT matching, multi-modal enrollment, and lightweight continual adaptation with 187k parameters.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.21789","ref_index":12,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Patch Hierarchical Attention Transformer for Efficient Particle Jet Tagging","primary_cat":"hep-ex","submitted_at":"2026-05-20T22:31:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"PHAT-JeT combines geometric message-passing with hierarchical patch attention to reach state-of-the-art accuracy and background rejection among resource-constrained jet tagging models on four benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.21627","ref_index":17,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Distribution-free root cause analysis","primary_cat":"stat.ME","submitted_at":"2026-05-20T18:40:04+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"CROC constructs finite-sample valid confidence sets for the root-cause index in multi-stream change detection using conformal p-values under independence and exchangeability assumptions.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.21421","ref_index":86,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"AIGaitor: Privacy-preserving and cloud-free motion analysis for everyone, using edge computing","primary_cat":"cs.CV","submitted_at":"2026-05-20T17:14:57+00:00","verdict":null,"verdict_confidence":null,"novelty_score":null,"formal_verification":null,"one_line_summary":null,"context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.21299","ref_index":68,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Tracing the ongoing emergence of human-like reasoning in Large Language Models","primary_cat":"cs.CL","submitted_at":"2026-05-20T15:28:52+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"LLMs function as accurate semantic processors for conditionals but do not replicate the pragmatic inferences that define human reasoning.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20723","ref_index":13,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Memory-Efficient Partitioned DNN Inference on Resource-Constrained Android Crowds","primary_cat":"cs.LG","submitted_at":"2026-05-20T05:21:54+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"CROWDio enables memory-efficient ONNX inference of DistilBERT on Android handsets by partitioning across devices with JIT loading, affinity scheduling, compressed transport and streaming, keeping per-device memory at 43 MB and cutting latency 34%.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20683","ref_index":35,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Layer-wise Token Compression for Efficient Document Reranking","primary_cat":"cs.IR","submitted_at":"2026-05-20T03:52:31+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Layer-wise Token Compression applies adaptive token pooling at middle transformer layers for cross-encoder rerankers, preserving MS MARCO ranking quality while raising QPS up to 25% on passages and 116% on documents, with added gains on listwise LLM rerankers and a regularizer effect for long inputs","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20423","ref_index":25,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"OSCToM: RL-Guided Adversarial Generation for High-Order Theory of Mind","primary_cat":"cs.AI","submitted_at":"2026-05-19T19:19:26+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"OSCToM uses RL-guided generation with an extended DSL and surrogate models to create nested belief conflict tasks, raising FANToM accuracy from 0.2% to 76% while being 6x more efficient.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20052","ref_index":16,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"PromptRad: Knowledge-Enhanced Multi-Label Prompt-Tuning for Low-Resource Radiology Report Labeling","primary_cat":"cs.CL","submitted_at":"2026-05-19T16:07:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"PromptRad reformulates multi-label radiology report classification as masked language modeling and enriches verbalizers with UMLS synonyms, outperforming baselines with only 32 training examples.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18752","ref_index":114,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Traditional statistical representations outperform generative AI in identifying expert peer reviewers","primary_cat":"cs.IR","submitted_at":"2026-05-18T17:59:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"TF-IDF identifies labeled experts in the top 25 recommendations 79.5% of the time versus 51.5% for GPT-4o mini on an astronomy observatory dataset.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18643","ref_index":34,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Post-Trained MoE Can Skip Half Experts via Self-Distillation","primary_cat":"cs.LG","submitted_at":"2026-05-18T16:50:48+00:00","verdict":null,"verdict_confidence":null,"novelty_score":null,"formal_verification":null,"one_line_summary":null,"context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18180","ref_index":39,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Canonical Regularisation of Wide Feature-Learning Neural Networks","primary_cat":"stat.ML","submitted_at":"2026-05-18T10:23:06+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"Derives geodesic ridge regularization and Riemannian Gibbs Process prior for feature-learning wide neural networks, generalizing kernel-regime results via function-space axiomatization.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18066","ref_index":68,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"TIDAL: Recovering Temporal Phase for Cloud Block Storage Placement from LLM-Derived Semantics","primary_cat":"cs.OS","submitted_at":"2026-05-18T08:49:16+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"TIDAL recovers temporal phase signals from LLM-derived semantics of provisioning metadata to enable complementary CVD placement, reducing overload frequency by 79.1% on production traces.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18007","ref_index":129,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Semantic Reranking at Inference Time for Hard Examples in Rhetorical Role Labeling","primary_cat":"cs.CL","submitted_at":"2026-05-18T08:03:02+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"RISE is an inference-time semantic reranking framework that refines low-confidence predictions in rhetorical role labeling using contrastively learned label representations, delivering an average +9.15 macro-F1 gain on hard examples across eight datasets and seven models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.17486","ref_index":73,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization","primary_cat":"cs.RO","submitted_at":"2026-05-17T14:55:32+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"DyGRO-VLA is a two-stage optimization framework for cross-task scaling of Vision-Language-Action models via dynamic grouped residual optimization in RL.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.17432","ref_index":27,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"DP-SelFT: Differentially Private Selective Fine-Tuning for Large Language Models","primary_cat":"cs.LG","submitted_at":"2026-05-17T12:55:11+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"DP-SelFT improves the privacy-utility trade-off for LLM fine-tuning by selecting robust layer subsets via DP synthetic data and perturbation-matched evaluation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.16991","ref_index":157,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Response-free item difficulty modelling for multiple-choice items with fine-tuned transformers: Component-wise representation and multi-task learning","primary_cat":"cs.CL","submitted_at":"2026-05-16T13:22:57+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Fine-tuned transformers with multi-task learning recover substantial wording-derived signal for item difficulty at small sample sizes typical in applied testing.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.15460","ref_index":32,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Differentially Private Motif-Preserving Multi-modal Hashing","primary_cat":"cs.IR","submitted_at":"2026-05-14T22:43:30+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"DMP-MH clips degrees to control triangle sensitivity, synthesizes an edge-DP graph with Noisy Mirror Descent, and distills it into dual-stream hash networks, beating private baselines by up to 11.4 mAP on MIRFlickr-25K and NUS-WIDE while keeping 92.5% of non-private performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.15413","ref_index":41,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Transformer Scalability Crisis: The First Comprehensive Empirical Analysis of Performance Walls in Modern Language Models","primary_cat":"cs.LG","submitted_at":"2026-05-14T20:57:15+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"Empirical tests on 118 transformers show success falling from 88.1% at 512 tokens to 0% at 2048 tokens, with compressed models achieving 649.2 tokens/sec/M parameters versus 12.5 for large generative ones.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.15299","ref_index":12,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning","primary_cat":"cs.IR","submitted_at":"2026-05-14T18:13:05+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Fortress stabilizes query-to-app relevance models by pruning features that cause inconsistent predictions across time periods while retaining predictive power from engagement signals.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.15104","ref_index":85,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents","primary_cat":"cs.CL","submitted_at":"2026-05-14T17:22:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A dataset-agnostic framework converts text tool-calling benchmarks to paired audio evaluations via TTS, speaker variation and noise, then evaluates seven omni-modal models showing model- and task-dependent performance with small text-to-voice gaps.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14260","ref_index":21,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"On the Burden of Achieving Fairness in Conformal Prediction","primary_cat":"stat.ML","submitted_at":"2026-05-14T02:02:06+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Pooled conformal calibration incurs irreducible group-wise coverage distortion scaled by cross-group quantile heterogeneity, with Equalized Coverage and Equalized Set Size in fundamental tension.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14257","ref_index":26,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Sakura at BEA 2026 Shared Task 1: What Makes Vocabulary Difficult?","primary_cat":"cs.CL","submitted_at":"2026-05-14T01:57:35+00:00","verdict":"ACCEPT","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Fine-tuned LLM and explainable models predict vocabulary difficulty with correlations r > 0.91 and r > 0.77, showing spelling difficulty and test item construction as key influences in addition to word production difficulty.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14075","ref_index":50,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Rethinking Layer Relevance in Large Language Models Beyond Cosine Similarity","primary_cat":"cs.LG","submitted_at":"2026-05-13T19:51:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Cosine similarity poorly predicts performance degradation from layer removal in LLMs, making direct accuracy-drop ablation a more reliable relevance metric.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14071","ref_index":32,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Distribution Corrected Offline Data Distillation for Large Language Models","primary_cat":"cs.CL","submitted_at":"2026-05-13T19:47:31+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A distribution-correction framework for offline LLM reasoning distillation improves accuracy on math benchmarks by adaptively aligning teacher supervision with the student's inference-time distribution.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"LLM post-training in domains such as math reasoning and safety alignment. The core idea behind this approach is to fine-tune a pretrained modelπθ to maximize accumulated rewards within a domain Qover self-generated trajectories: JRL(θ) =E x∼Q, y∼π θ(·|x) [R(y, x)],(1) where R(x, y) is a reward function that scores a response y (e.g., a reasoning trajectory). Following the policy gradient principle [32], we derive the following gradient update for Eq. (1): ∇θJRL(θ) =E x∼Q,y∼π θ(· |x)| {z } On-policy sampling  R(x, y)· |y|X t=0 logπ θ(yt |x, y :