INDEX
Every paper Pith has read and judged, in one searchable index.
-
cs.CV arXiv submitted 2026-07-14VLMs gain 28 points on cross-image comparison with triplet rewards
Lin Peng +4 · “CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models”
2607.12786 -
cs.CV arXiv submitted 2026-07-14Diffusion-guided Gaussians clean up unseen endoscopic anatomy
Cheng-Tai Hsieh +8 · “ExtraGS: Enhancing Endoscopic View Extrapolation via Diffusion-Guided 3D Gaussian Splatting”
2607.12785 -
cs.RO arXiv submitted 2026-07-14Safety layer only kicks in when robots head toward danger
Paolo Magliano +4 · “Directional Constraints for Efficient Exploration in Safe Reinforcement Learning”
2607.12784 -
cs.CV arXiv submitted 2026-07-14Keep every spectral band, train only ~2% of the model
Mingzhen Xu +11 · “MBTI: A Multi-Branch Efficient Fine-Tuning Framework for Hyperspectral Image Classification with Foundation Models”
2607.12782 -
quant-ph arXiv submitted 2026-07-14Transformers hit a length wall on exact quantum circuits
Mehdi Saeedi +2 · “When Close Enough Is Not Enough: Autoregressive Drift in Quantum Circuit Synthesis”
2607.12780 -
math.OC arXiv submitted 2026-07-14Learning ADMM makes scenario MPC fast enough for real time
Trinh Tran +2 · “Learning-enabled Acceleration of Scenario-based Model Predictive Control”
2607.12775 -
cs.CV arXiv submitted 2026-07-14Frozen light models beat heavy ABAW baselines with calibration
Aleksei Bakin +1 · “HSEmotion Team at the 11th ABAW Challenge: Multi-Task Learning and Ambivalence/Hesitancy Video Recognition”
2607.12774 -
cs.LG arXiv submitted 2026-07-1430B LLM beats specialized model on exact reaction pathways
Xingyu Dang +3 · “Learning Mechanistic Reasoning for Chemical Reactions with Large Language Models”
2607.12771 -
cs.AI arXiv submitted 2026-07-14Bayesian score ends length bias in multiple-choice LM tests
Koen Oostermeijer · “Accuracy and Normalized Accuracy under Length Bias: Analysis, Guidelines, and a Bayesian Alternative”
2607.12767 -
cs.CR arXiv submitted 2026-07-14Cloud-orchestrated QKD network spans three vendors and non-QKD domains
Konstantinos Krilakis +3 · “A Scalable Cloud-Orchestrated and Service-Oriented Multi-Domain QKD Network with PQC Integration”
2607.12765 -
cs.CV arXiv submitted 2026-07-14Agent rewrites its own knowledge hypergraph mid-reasoning
Jiashi Lin +14 · “EvoGraph-R1: Self-Evolving Multimodal Knowledge Hypergraphs for Agentic Retrieval”
2607.12764 -
cs.LG arXiv submitted 2026-07-14Penalty weights make federated RL microgrids safer than FedAvg
Usman Haider +1 · “Constraint-Aware Aggregation for Federated Reinforcement Learning in Microgrid Energy Coordination”
2607.12763 -
math.NA arXiv submitted 2026-07-14Mesh-free solver keeps every phase area exact in multi-phase flow
Tokuhiro Eto · “A Structure-Preserving Method of Fundamental Solutions for the Multi-Phase Mullins-Sekerka Flow”
2607.12759 -
cs.CV arXiv submitted 2026-07-14One VLM compresses its own visual tokens into a few memory slots
Yupeng Zheng +3 · “VisCo: Leveraging Large Language Models as Intrinsic Encoders for Visual Token Compression”
2607.12756 -
cs.HC arXiv submitted 2026-07-14Judgment and virtue can blunt the risks of opaque AI
Nathan G. Wood +1 · “Practical Judgment, Virtue, and Intuition in the Use of Opaque AI-Enabled Systems”
2607.12755 -
cs.CV arXiv submitted 2026-07-14Flow starts from the low-quality image
Shuwei Huang +4 · “RFMSR: Residual Flow Matching for Image Super-Resolution”
2607.12753 -
cs.CV arXiv submitted 2026-07-14LMMs detect and fix 3D/4D hallucinations without retraining
Hongbo Wang +5 · “Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation”
2607.12752 -
cs.CV arXiv submitted 2026-07-14Public dataset grades colorectal cancer at four magnifications
Elham Amjadi +6 · “CRC-HGD: A Histopathological Image Dataset for Grading Colorectal Cancer”
2607.12750 -
cs.CV arXiv submitted 2026-07-14Satellite tiles shrink to 71 ranked dairy-farm candidates
Usman Haider +2 · “Weakly Supervised Spatio-Temporal Candidate Discovery of Dairy Farm Sites from Seasonal Satellite Imagery”
2607.12748 -
cs.AI arXiv submitted 2026-07-14Train on 100 successes, then flag the steps that break agent runs
Samuel Yeh +3 · “Tracing Agentic Failure from the Flow of Success”
2607.12747 -
cs.CV arXiv submitted 2026-07-14Coupled camera-display map doubles color fidelity
Ruikang Li +5 · “Color Pass-Through via Camera-Display Coupling”
2607.12746 -
cs.CR arXiv submitted 2026-07-14Graph fragility sets how often encrypted fleets must re-key
Sai Sandeep Damera +1 · “Stability Buys Time: A Re-Keying Game for Encrypted Multi-Agent Control”
2607.12742 -
cs.CL arXiv submitted 2026-07-14Stance flexibility under prompts is orthogonal to model capability
Binwen Liu +1 · “Epistemic Stance Flexibility Probing: Measuring Prompt-Conditioned Register Shift in Large Language Models”
2607.12739 -
cs.HC arXiv submitted 2026-07-14AI research assistants should serve teams
Diya Mirji +11 · “A\"ira: Rethinking AI Research Assistants for Interdisciplinary Science”
2607.12736 -
cs.LG arXiv submitted 2026-07-14Label-free prior speeds grokking 2.7× more reliably than labels
Gunner Levi Howe · “What Makes a Representational Prior Work? Feature Families, Label-Free Invariances, and Critical Windows in Grokking”
2607.12735 -
cs.AI arXiv submitted 2026-07-14LLMs spot mutated formal systems but miss the rule changes
Julius Steiglechner +2 · “LLMs Can See the Smoke but not the Fire: Evaluating Abductive Reasoning with Elenchos”
2607.12733 -
cs.RO arXiv submitted 2026-07-14Every particle stays feasible
Jiayun Li +1 · “Globalized Constrained Stein Variational Inference for Diverse Feasible Robot Motion Planning”
2607.12732 -
cs.LG arXiv submitted 2026-07-14Uncertainty placement is backbone-dependent once inputs are reconstructed
Sarah Al-Shareeda +2 · “Learning-based Probabilistic Load Forecasting with Post-hoc and In-model Uncertainty”
2607.12730 -
hep-ph arXiv submitted 2026-07-14XGBoost maps non-Gaussian HEP likelihoods without losing physics
Jorge Alda +3 · “Physically Consistent Parameter Inference: Transparent Machine Learning Emulation in High Energy Physics and Cosmology”
2607.12726 -
cs.SD arXiv submitted 2026-07-14Codec tokens morph source rhythm with palette timbre
Emmanouil Karystinaios · “Neural Morphing: Sequence-Optimized Token-Level Morphing in Neural Audio Codecs”
2607.12725 -
cs.CR arXiv submitted 2026-07-14LLMs plus formal checks auto-find and patch container path escapes
Qiyuan Fan +5 · “Bulkhead: Automated Semantic Detection and Remediation of Container Escape Vulnerabilities”
2607.12723 -
cs.NI arXiv submitted 2026-07-14Hybrid FA-PSO recovers wall material parameters near the CRLB
Zhuowei Li +9 · “High-Precision Hybrid FA-PSO Based Inversion of Building Material Parameters for Fundamental Wireless Performance Evaluation”
2607.12721 -
cs.DB arXiv submitted 2026-07-14Tables retrieved as joinable graphs
Daomin Ji +4 · “GRAFT: Graph-Matched Retrieval and Fusion of Tables in Data Lakes”
2607.12717 -
cs.CR arXiv submitted 2026-07-14USB dongle encrypts whole disks
Enrique Soriano-Salvador +1 · “Auditable and Transparent Fully Authenticated Disk Encryption via USB Storage Interposition”
2607.12716 -
math.NA arXiv submitted 2026-07-14Mass-conserving dynamics turn saddles into stable states
Longjing Li +2 · “Mass-Conserving Saddle Dynamics via Generalized Inner Product: Theory, Algorithms, and Applications”
2607.12715 -
cs.IR arXiv submitted 2026-07-14Purchase as interest exit cuts post-buy redundancy by over 60%
Yipin Dai +6 · “Learning to Forget: Satiation-Aware Long-Sequence Transducers for Mitigating Post-Purchase Redundancy”
2607.12714 -
cs.SE arXiv submitted 2026-07-14Line-anchored feedback cuts AI edit tokens up to 58%
William Franz Lamberti · “Line-Anchored Feedback Cuts Token Costs and Improves Correctness in AI Code Editing”
2607.12713 -
cs.AI arXiv submitted 2026-07-14MaxSAT feedback lifts VLM Sudoku solves by keeping only consistent cells
Pedro Orvalho +2 · “MaxSAT-Based Feedback for Guiding Vision-Language Models in Sudoku”
2607.12711 -
cs.IT arXiv submitted 2026-07-14Superimposed symbols raise peak SE in hybrid massive MIMO
Wei Jiang +1 · “Superimposed Transmission for Cooperative Cellular and Cell-Free Massive MIMO Systems”
2607.12709 -
cs.SD arXiv submitted 2026-07-14Edit one speech style category
Haowei Lou +6 · “AutoSIFT: Automatic Style Sifting for Controllable Speech Generation with Arbitrary Style Infilling”
2607.12706 -
cs.CV arXiv submitted 2026-07-14Per-label style mixing lifts multi-label aerial DG by 5 mAP
Alaa Almouradi +1 · “Label-Decoupled Style Augmentation for Domain Generalization in Multi-Label Remote Sensing Scene Classification”
2607.12704 -
cs.RO arXiv submitted 2026-07-14Humanoid learns soccer dribbling from depth images alone
Flavio Maiorana +7 · “Vision-Based Dribbling for Humanoid Soccer via Privileged Representation Learning”
2607.12702 -
eess.SY arXiv submitted 2026-07-14Converter power and grid strength jointly set VSC stability
Alessandra Casiraghi (1) +5 · “Stability Analysis of Grid-Following and Grid-Forming Converters Connected to Generators”
2607.12697 -
cs.CL arXiv submitted 2026-07-14Cost-aware drafts cut MoE expert thrashing
Jincheng Xie +6 · “Less Experts, Faster Decoding: Cost-Aware Speculative Decoding for Mixture-of-Experts”
2607.12696 -
cs.IT arXiv submitted 2026-07-14THz pinching antennas beat phased arrays via proximity
Wei Jiang +1 · “Pinching-Antenna-Assisted Terahertz Communications: Modeling and Benchmarking”
2607.12695 -
quant-ph arXiv submitted 2026-07-14Eleven quantum heat solvers ranked under one fixed contract
Mahmoud Elkarargy +3 · “Quantum PDE Solvers in Practice: Application-Driven Benchmarking of the Heat Equation”
2607.12688 -
cs.CL arXiv submitted 2026-07-14PPO critic becomes the confidence signal for language number prediction
Mehak Dhaliwal +4 · “From Critic to Confidence: PPO for Language-Based Quantitative Prediction with Confidence Estimation”
2607.12687 -
cs.CL arXiv submitted 2026-07-14Segregate then fuse: SeRIn hits SOTA multimodal sentiment scores
Alexios Filippakopoulos +4 · “Segregate, Refine, Integrate: Decomposing Multimodal Fusion for Sentiment Analysis”
2607.12686 -
cs.CV arXiv submitted 2026-07-14Brain-only intensity fix lifts nnU-Net TBI lesion Dice to 0.63
Inhwa Son +3 · “Lesion Segmentation in Moderate to Severe Traumatic Brain Injury: An nnU-Net Based Approach with Adaptive Normalization in the AIMS-TBI 2025 Challenge”
2607.12684 -
cs.CV arXiv submitted 2026-07-14Mamba swap cuts YOLO26 FLOPs 12% with near-zero accuracy loss
Sheng-Wei Chan +6 · “MambaPSA: A Mamba-based Replacement for C2PSA in YOLO26”
2607.12681