AI Research News Feeds for February 7th, 2026

AI RESEARCH PAPERS & ACADEMIC SOURCES

• Log2Motion: Biomechanical Motion Synthesis from Touch Logs
• STELLAR: Structure-guided LLM Assertion Retrieval and Generation for Formal Verification
• CGRA4ML: A Hardware/Software Framework to Implement Neural Networks for Scientific Edge Computing
• KANFIS: A Neuro-Symbolic Framework for Interpretable and Uncertainty-Aware Learning
• Learning to Share: Selective Memory for Efficient Parallel Agentic Systems
• Exploring Silicon-Based Societies: An Early Study of the Moltbook Agent Community
• Quality Model for Machine Learning Components
• Bypassing AI Control Protocols via Agent-as-a-Proxy Attacks
• Large-Ensemble Simulations Reveal Links Between Atmospheric Blocking Frequency and Sea Surface Temperature Variability
• ARCHI-TTS: A flow-matching-based Text-to-Speech Model with Self-supervised Semantic Aligner and Accelerated Inference
• Semantic Search over 9 Million Mathematical Theorems
• EGSS: Entropy-guided Stepwise Scaling for Reliable Software Engineering
• Spider-Sense: Intrinsic Risk Sensing for Efficient Agent Defense with Hierarchical Adaptive Screening
• Benchmarking Affordance Generalization with BusyBox
• Ontology-Driven Robotic Specification Synthesis
• LMMRec: LLM-driven Motivation-aware Multimodal Recommendation
• Sovereign-by-Design A Reference Architecture for AI and Blockchain Enabled Systems
• Quantum Reinforcement Learning with Transformers for the Capacitated Vehicle Routing Problem
• DECO: Decoupled Multimodal Diffusion Transformer for Bimanual Dexterous Manipulation with a Plugin Tactile Adapter
• Capture the Flags: Family-Based Evaluation of Agentic LLMs via Semantics-Preserving Transformations
• AI Agent Systems for Supply Chains: Structured Decision Prompts and Memory Retrieval
• HyperPotter: Spell the Charm of High-Order Interactions in Audio Deepfake Detection
• Exploring AI-Augmented Sensemaking of Patient-Generated Health Data: A Mixed-Method Study with Healthcare Professionals in Cardiac Risk Reduction
• Towards Green AI: Decoding the Energy of LLM Inference in Software Development
• TimelyFreeze: Adaptive Parameter Freezing Mechanism for Pipeline Parallelism
• Automated Customization of LLMs for Enterprise Code Repositories Using Semantic Scopes
• FHAIM: Fully Homomorphic AIM For Private Synthetic Data Generation
• Metric Hedonic Games on the Line
• Compound Deception in Elite Peer Review: A Failure Mode Taxonomy of 100 Fabricated Citations at NeurIPS 2025
• Evaluating the impact of word embeddings on similarity scoring in practical information retrieval
• Emulating Aggregate Human Choice Behavior and Biases with GPT Conversational Agents
• Reactive Knowledge Representation and Asynchronous Reasoning
• Graph-based Agent Memory: Taxonomy, Techniques, and Applications
• Determining Energy Efficiency Sweet Spots in Production LLM Inference
• Nonlinearity as Rank: Generative Low-Rank Adapter with Radial Basis Functions
• Anchored Policy Optimization: Mitigating Exploration Collapse Via Support-Constrained Rectification
• Mitigating Hallucination in Financial Retrieval-Augmented Generation via Fine-Grained Knowledge Verification
• LeakBoost: Perceptual-Loss-Based Membership Inference Attack
• RL-VLA$^3$: Reinforcement Learning VLA Accelerating via Full Asynchronism
• NEX: Neuron Explore-Exploit Scoring for Label-Free Chain-of-Thought Selection and Model Ranking
• STProtein: predicting spatial protein expression from multi-omics data
• TKG-Thinker: Towards Dynamic Reasoning over Temporal Knowledge Graphs via Agentic Reinforcement Learning
• OmniVideo-R1: Reinforcing Audio-visual Reasoning with Query Intention and Modality Attention
• BABE: Biology Arena BEnchmark
• Beyond Manual Planning: Seating Allocation for Large Organizations
• Agent2Agent Threats in Safety-Critical LLM Assistants: A Human-Centric Taxonomy
• A Guide to Large Language Models in Modeling and Simulation: From Core Techniques to Critical Challenges
• Enabling Automatic Disordered Speech Recognition: An Impaired Speech Dataset in the Akan Language
• Geographically-aware Transformer-based Traffic Forecasting for Urban Motorway Digital Twins
• Learning Event-Based Shooter Models from Virtual Reality Experiments
• DyTopo: Dynamic Topology Routing for Multi-Agent Reasoning via Semantic Matching
• Doc2Spec: Synthesizing Formal Programming Specifications from Natural Language via Grammar Induction
• Extracting Recurring Vulnerabilities from Black-Box LLM-Generated Software
• Steering Externalities: Benign Activation Steering Unintentionally Increases Jailbreak Risk for Large Language Models
• Semantic-level Backdoor Attack against Text-to-Image Diffusion Models
• Phantom Transfer: Data-level Defences are Insufficient Against Data Poisoning
• Evaluating Kubernetes Performance for GenAI Inference: From Automatic Speech Recognition to LLM Summarization
• PriMod4AI: Lifecycle-Aware Privacy Threat Modeling for AI Systems using LLM
• Attack Selection Reduces Safety in Concentrated AI Control Settings against Trusted Monitoring
• ASA: Activation Steering for Tool-Calling Domain Adaptation
• Graph--Theoretic Analysis of Phase Optimization Complexity in Variational Wave Functions for Heisenberg Antiferromagnets
• From Fragmentation to Integration: Exploring the Design Space of AI Agents for Human-as-the-Unit Privacy Management
• Do Vision-Language Models Respect Contextual Integrity in Location Disclosure?
• AudioSAE: Towards Understanding of Audio-Processing Models with Sparse AutoEncoders
• Artificial Intelligence as Strange Intelligence: Against Linear Models of Intelligence
• MINT: Minimal Information Neuro-Symbolic Tree for Objective-Driven Knowledge-Gap Reasoning and Active Elicitation
• Evaluating Large Language Models on Solved and Unsolved Problems in Graph Theory: Implications for Computing Education
• Towards Reducible Uncertainty Modeling for Reliable Large Language Model Agents
• VERA-MH: Reliability and Validity of an Open-Source AI Safety Evaluation in Mental Health
• GAMMS: Graph based Adversarial Multiagent Modeling Simulator
• Understanding LLM Evaluator Behavior: A Structured Multi-Evaluator Framework for Merchant Risk Assessment
• HugRAG: Hierarchical Causal Knowledge Graph Design for RAG
• First Proof
• Traceable Cross-Source RAG for Chinese Tibetan Medicine Question Answering
• Surgery: Mitigating Harmful Fine-Tuning for Large Language Models via Attention Sink
• Automatic Cognitive Task Generation for In-Situ Evaluation of Embodied Agents
• Beyond Cosine Similarity
• Hallucination-Resistant Security Planning with a Large Language Model
• Position: Universal Time Series Foundation Models Rest on a Category Error
• Aspect-Aware MOOC Recommendation in a Heterogeneous Network
• PieArena: Frontier Language Agents Achieve MBA-Level Negotiation Performance and Reveal Novel Behavioral Differences
• ProAct: Agentic Lookahead in Interactive Environments
• PATHWAYS: Evaluating Investigation and Context Discovery in AI Web Agents
• RaBiT: Residual-Aware Binarization Training for Accurate and Efficient LLMs
• Clinical Validation of Medical-based Large Language Model Chatbots on Ophthalmic Patient Queries with LLM-based Evaluation
• Advancing Opinion Dynamics Modeling with Neural Diffusion-Convection-Reaction Equation
• THOR: Inductive Link Prediction over Hyper-Relational Knowledge Graphs
• TangramSR: Can Vision-Language Models Reason in Continuous Geometric Space?
• Reasoning-guided Collaborative Filtering with Language Models for Explainable Recommendation
• Conditional Diffusion Guidance under Hard Constraint: A Stochastic Analysis Approach
• SDFP: Speculative Decoding with FIT-Pruned Models for Training-Free and Plug-and-Play LLM Acceleration
• Phi-Former: A Pairwise Hierarchical Approach for Compound-Protein Interactions Prediction
• ALIVE: Awakening LLM Reasoning via Adversarial Learning and Instructive Verbal Evaluation
• Explainable AI: A Combined XAI Framework for Explaining Brain Tumour Detection Models