ProbBrain - AI News Terminal https://probbrain.com/news AI lab announcements, arXiv papers, GitHub and HuggingFace trending, AI Hacker News, and independent writers - aggregated, deduped, and posted to @ProbBrain on Telegram. en Mon, 04 May 2026 12:23:31 +0000 15 [GitHub Trending] sponsors/msitarzewski — A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables. https://github.com/sponsors/msitarzewski github|https://github.com/sponsors/msitarzewski Mon, 04 May 2026 10:51:20 +0000 code [GitHub Trending] virattt/dexter — An autonomous agent for deep financial research https://github.com/virattt/dexter github|https://github.com/virattt/dexter Mon, 04 May 2026 10:51:20 +0000 code [GitHub Trending] fspecii/ace-step-ui — 🎵 The Ultimate Open Source Suno Alternative - Professional UI for ACE-Step 1.5 AI Music Generation. Free, local, unlimited. Stop paying for Suno! https://github.com/fspecii/ace-step-ui github|https://github.com/fspecii/ace-step-ui Mon, 04 May 2026 10:51:20 +0000 code [GitHub Trending] cocoindex-io/cocoindex — Incremental engine for long horizon agents 🌟 Star if you like it! https://github.com/cocoindex-io/cocoindex github|https://github.com/cocoindex-io/cocoindex Mon, 04 May 2026 10:51:20 +0000 code [arXiv cs.LG] Cloud Is Closer Than It Appears: Revisiting the Tradeoffs of Distributed Real-Time Inference https://arxiv.org/abs/2605.00005 arxiv-cs.LG|https://arxiv.org/abs/2605.00005 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] FedACT: Concurrent Federated Intelligence across Heterogeneous Data Sources https://arxiv.org/abs/2605.00011 arxiv-cs.LG|https://arxiv.org/abs/2605.00011 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] What Physics do Data-Driven MoCap-to-Radar Models Learn? https://arxiv.org/abs/2605.00018 arxiv-cs.LG|https://arxiv.org/abs/2605.00018 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] AirFM-DDA: Air-Interface Foundation Model in the Delay-Doppler-Angle Domain for AI-Native 6G https://arxiv.org/abs/2605.00020 arxiv-cs.LG|https://arxiv.org/abs/2605.00020 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Learning physically grounded traffic accident reconstruction from public accident reports https://arxiv.org/abs/2605.00050 arxiv-cs.LG|https://arxiv.org/abs/2605.00050 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Smart Ensemble Learning Framework for Predicting Groundwater Heavy Metal Pollution https://arxiv.org/abs/2605.00056 arxiv-cs.LG|https://arxiv.org/abs/2605.00056 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Information-Theoretic Generalization Bounds for Stochastic Gradient Descent with Predictable Virtual Noise https://arxiv.org/abs/2605.00064 arxiv-cs.LG|https://arxiv.org/abs/2605.00064 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Human-in-the-Loop Meta Bayesian Optimization for Fusion Energy and Scientific Applications https://arxiv.org/abs/2605.00068 arxiv-cs.LG|https://arxiv.org/abs/2605.00068 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Soft-MSM: Differentiable Context-Aware Elastic Alignment for Time Series https://arxiv.org/abs/2605.00069 arxiv-cs.LG|https://arxiv.org/abs/2605.00069 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] CRADIPOR: Crash Dispersion Predictor https://arxiv.org/abs/2605.00070 arxiv-cs.LG|https://arxiv.org/abs/2605.00070 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Hyperspherical Forward-Forward with Prototypical Representations https://arxiv.org/abs/2605.00082 arxiv-cs.LG|https://arxiv.org/abs/2605.00082 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Comparative Analysis of Polygon-Based and Global Machine Learning Models for Bus Occupancy Prediction https://arxiv.org/abs/2605.00083 arxiv-cs.LG|https://arxiv.org/abs/2605.00083 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] SPLICE: Latent Diffusion over JEPA Embeddings for Conformal Time-Series Inpainting https://arxiv.org/abs/2605.00126 arxiv-cs.LG|https://arxiv.org/abs/2605.00126 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization https://arxiv.org/abs/2605.00130 arxiv-cs.LG|https://arxiv.org/abs/2605.00130 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Smart Profit-Aware Crop Advisory System: Kisan AI https://arxiv.org/abs/2605.00133 arxiv-cs.LG|https://arxiv.org/abs/2605.00133 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Technical Report: Activation Residual Hessian Quantization (ARHQ) for Low-Bit LLM Quantization https://arxiv.org/abs/2605.00140 arxiv-cs.LG|https://arxiv.org/abs/2605.00140 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Wasserstein Distributionally Robust Regret Optimization for Reinforcement Learning from Human Feedback https://arxiv.org/abs/2605.00155 arxiv-cs.LG|https://arxiv.org/abs/2605.00155 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Consistent Diffusion Language Models https://arxiv.org/abs/2605.00161 arxiv-cs.LG|https://arxiv.org/abs/2605.00161 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Towards A Generative Protein Evolution Machine with DPLM-Evo https://arxiv.org/abs/2605.00182 arxiv-cs.LG|https://arxiv.org/abs/2605.00182 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Introducing WARM-VR: Benchmark Dataset for Multimodal Wearable Affect Recognition in Virtual Reality https://arxiv.org/abs/2605.00184 arxiv-cs.LG|https://arxiv.org/abs/2605.00184 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Fair Dataset Distillation via Cross-Group Barycenter Alignment https://arxiv.org/abs/2605.00185 arxiv-cs.LG|https://arxiv.org/abs/2605.00185 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] OTSS: Output-Targeted Soft Segmentation for Contextual Decision-Weight Learning https://arxiv.org/abs/2605.00193 arxiv-cs.LG|https://arxiv.org/abs/2605.00193 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Diversity in Large Language Models under Supervised Fine-Tuning https://arxiv.org/abs/2605.00195 arxiv-cs.LG|https://arxiv.org/abs/2605.00195 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] State Stream Transformer (SST) V2: Parallel Training of Nonlinear Recurrence for Latent Space Reasoning https://arxiv.org/abs/2605.00206 arxiv-cs.LG|https://arxiv.org/abs/2605.00206 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] CompleteRXN: Toward Completing Open Chemical Reaction Databases https://arxiv.org/abs/2605.00222 arxiv-cs.LG|https://arxiv.org/abs/2605.00222 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Bayesian Optimization in Linear Time https://arxiv.org/abs/2605.00237 arxiv-cs.LG|https://arxiv.org/abs/2605.00237 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] NLPOpt-Net: A Learning Method for Nonlinear Optimization with Feasibility Guarantees https://arxiv.org/abs/2605.00260 arxiv-cs.LG|https://arxiv.org/abs/2605.00260 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Pessimism-Free Offline Learning in General-Sum Games via KL Regularization https://arxiv.org/abs/2605.00264 arxiv-cs.LG|https://arxiv.org/abs/2605.00264 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Polaris: Coupled Orbital Polar Embeddings for Hierarchical Concept Learning https://arxiv.org/abs/2605.00265 arxiv-cs.LG|https://arxiv.org/abs/2605.00265 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.LG] Jailbroken Frontier Models Retain Their Capabilities https://arxiv.org/abs/2605.00267 arxiv-cs.LG|https://arxiv.org/abs/2605.00267 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Putting HUMANS first: Efficient LAM Evaluation with Human Preference Alignment https://arxiv.org/abs/2605.00022 arxiv-cs.CL|https://arxiv.org/abs/2605.00022 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] NorBERTo: A ModernBERT Model Trained for Portuguese with 331 Billion Tokens Corpus https://arxiv.org/abs/2605.00086 arxiv-cs.CL|https://arxiv.org/abs/2605.00086 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] How Frontier LLMs Adapt to Neurodivergence Context: A Measurement Framework for Surface vs. Structural Change in System-Prompted Responses https://arxiv.org/abs/2605.00113 arxiv-cs.CL|https://arxiv.org/abs/2605.00113 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] ViLegalNLI: Natural Language Inference for Vietnamese Legal Texts https://arxiv.org/abs/2605.00116 arxiv-cs.CL|https://arxiv.org/abs/2605.00116 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Cultural Benchmarking of LLMs in Standard and Dialectal Arabic Dialogues https://arxiv.org/abs/2605.00119 arxiv-cs.CL|https://arxiv.org/abs/2605.00119 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Timing is Everything: Temporal Scaffolding of Semantic Surprise in Humor https://arxiv.org/abs/2605.00143 arxiv-cs.CL|https://arxiv.org/abs/2605.00143 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] RSAT: Structured Attribution Makes Small Language Models Faithful Table Reasoners https://arxiv.org/abs/2605.00199 arxiv-cs.CL|https://arxiv.org/abs/2605.00199 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Confidence Estimation in Automatic Short Answer Grading with LLMs https://arxiv.org/abs/2605.00200 arxiv-cs.CL|https://arxiv.org/abs/2605.00200 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Why Do LLMs Struggle in Strategic Play? Broken Links Between Observations, Beliefs, and Actions https://arxiv.org/abs/2605.00226 arxiv-cs.CL|https://arxiv.org/abs/2605.00226 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Persona-Grounded Safety Evaluation of AI Companions in Multi-Turn Conversations https://arxiv.org/abs/2605.00227 arxiv-cs.CL|https://arxiv.org/abs/2605.00227 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Estimating LLM Grading Ability and Response Difficulty in Automatic Short Answer Grading via Item Response Theory https://arxiv.org/abs/2605.00238 arxiv-cs.CL|https://arxiv.org/abs/2605.00238 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Lost in State Space: Probing Frozen Mamba Representations https://arxiv.org/abs/2605.00253 arxiv-cs.CL|https://arxiv.org/abs/2605.00253 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Retrieval-Augmented Reasoning for Chartered Accountancy https://arxiv.org/abs/2605.00257 arxiv-cs.CL|https://arxiv.org/abs/2605.00257 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] How Language Models Process Out-of-Distribution Inputs: A Two-Pathway Framework https://arxiv.org/abs/2605.00269 arxiv-cs.CL|https://arxiv.org/abs/2605.00269 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Are You the A-hole? A Fair, Multi-Perspective Ethical Reasoning Framework https://arxiv.org/abs/2605.00270 arxiv-cs.CL|https://arxiv.org/abs/2605.00270 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] What Don't You Understand? Using Large Language Models to Identify and Characterize Student Misconceptions About Challenging Topics https://arxiv.org/abs/2605.00294 arxiv-cs.CL|https://arxiv.org/abs/2605.00294 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Structure-Aware Chunking for Tabular Data in Retrieval-Augmented Generation https://arxiv.org/abs/2605.00318 arxiv-cs.CL|https://arxiv.org/abs/2605.00318 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Prompt-Induced Score Variance in Zero-Shot Binary Vision-Language Safety Classification https://arxiv.org/abs/2605.00326 arxiv-cs.CL|https://arxiv.org/abs/2605.00326 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Budget-Aware Routing for Long Clinical Text https://arxiv.org/abs/2605.00336 arxiv-cs.CL|https://arxiv.org/abs/2605.00336 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Making Every Verified Token Count: Adaptive Verification for MoE Speculative Decoding https://arxiv.org/abs/2605.00342 arxiv-cs.CL|https://arxiv.org/abs/2605.00342 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents https://arxiv.org/abs/2605.00356 arxiv-cs.CL|https://arxiv.org/abs/2605.00356 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] From Backward Spreading to Forward Replay: Revisiting Target Construction in LLM Parameter Editing https://arxiv.org/abs/2605.00358 arxiv-cs.CL|https://arxiv.org/abs/2605.00358 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Unlearning What Matters: Token-Level Attribution for Precise Language Model Unlearning https://arxiv.org/abs/2605.00364 arxiv-cs.CL|https://arxiv.org/abs/2605.00364 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Language-free Experience at Expo 2025 Osaka https://arxiv.org/abs/2605.00373 arxiv-cs.CL|https://arxiv.org/abs/2605.00373 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Agentic AI for Substance Use Education: Integrating Regulatory and Scientific Knowledge Sources https://arxiv.org/abs/2605.00383 arxiv-cs.CL|https://arxiv.org/abs/2605.00383 Mon, 04 May 2026 04:00:00 +0000 paper [arXiv cs.CL] Agent Capsules: Quality-Gated Granularity Control for Multi-Agent LLM Pipelines https://arxiv.org/abs/2605.00410 arxiv-cs.CL|https://arxiv.org/abs/2605.00410 Mon, 04 May 2026 04:00:00 +0000 paper