AI × Crypto News Feed

Dsat: A Native SAT Solver for Discrete Logic

Researchers introduce DSAT, a native SAT solver designed to work directly with discrete variables rather than converting them to binary Boolean variables. The solver applies traditional SAT techniques like unit resolution and clause learning to discrete logic, offering potential computational and semantic advantages over existing binarization approaches for applications in probabilistic reasoning, planning, and explainable AI.

Do Self-Evolving Agents Forget? Capability Degradation and Preservation in Lifelong LLM Agent Adaptation

Researchers identify capability erosion in self-evolving LLM agents, where systems adapting to new tasks progressively lose previously learned abilities across workflow, skill, model, and memory dimensions. The study proposes Capability-Preserving Evolution (CPE), a stabilization framework that maintains performance on existing tasks while enabling new adaptations, demonstrating improvements in retained capability stability across all evolution channels.

Beyond Continuity: Challenges of Context Switching in Multi-Turn Dialogue with LLMs

Researchers tested how well Large Language Models handle multi-turn conversations with topic shifts, finding that most LLMs struggle to detect when users pivot to new topics and incorrectly carry over irrelevant context from previous exchanges. The study reveals that only advanced reasoning models and strongly instructed LLMs perform accurately, while open-weight models frequently fail even with explicit cues, highlighting a critical robustness gap in production LLM deployments.

Revisiting Mixture Policies in Entropy-Regularized Actor-Critic

Researchers propose a marginalized reparameterization (MRP) estimator to enable practical use of mixture policies in reinforcement learning, addressing a long-standing gap between theoretical potential and practical implementation. By reducing variance compared to likelihood-ratio methods, MRP mixture policies achieve performance parity with standard Gaussian policies while offering greater flexibility in continuous action spaces.

Continuity Laws for Sequential Models

Researchers formalize the concept of model continuity in sequential neural networks, finding that S4 maintains stable continuous behavior while Mamba's S6 exhibits sensitivity to input amplitude despite continuous-time origins. The study establishes empirical alignment between task continuity, model continuity, and performance, with practical implications for temporal subsampling strategies.

Behavioral Determinants of Deployed AI Agents in Social Networks: A Multi-Factor Study of Personality, Model, and Guardrail Specification

Researchers deployed thirteen AI agents on Moltbook, a Reddit-like social network for AI systems, to study how configuration specifications affect emergent social behavior. Results show personality specification is the dominant factor influencing agent responses, while underlying LLM models and operational rules have more moderate effects on communication style and topic engagement.

AI Native Asset Intelligence

Researchers introduce AI-native asset intelligence, a framework that structures fragmented security data across cloud environments to enable consistent, contextual prioritization of cybersecurity threats. The system combines asset modeling with intelligent scoring mechanisms that separate intrinsic exposure from business context, tested on 131,625 production resources across 15 vendors.

Internal vs. External: Comparing Deliberation and Evolution for Multi-Agent Constitutional Design

Researchers conducted the first controlled comparison of internal deliberation versus external evolution for designing behavioral rules in multi-agent AI systems across three social environments. Evolution significantly outperformed deliberation in collective-action settings, but both methods failed to improve outcomes in bilateral trading, with evolution's advantage reversing under certain economic conditions where it enforced value-destroying cooperation.

Personalized Alignment Revisited: The Necessity and Sufficiency of User Diversity

This theoretical computer science paper establishes formal conditions for efficient personalized alignment in large language models, proving that user diversity—specifically whether user-specific parameters span latent reward directions—is both necessary and sufficient for optimal statistical efficiency. The research provides rigorous mathematical foundations for adapting AI systems to heterogeneous user preferences.

Contextual Plackett-Luce: An Efficient Neural Model for Probabilistic Sequence Selection under Ambiguity

Researchers propose Contextual Plackett-Luce (CPL), a neural probabilistic model for sequence selection that balances computational efficiency with representational flexibility. The model addresses the challenge of predicting multi-modal outputs from single training examples by combining parallel scoring with lightweight autoregressive selection, demonstrating improvements on path prediction and subset selection tasks.

From Traditional Taggers to LLMs: A Comparative Study of POS Tagging for Medieval Romance Languages

Researchers conducted a systematic evaluation of large language models for part-of-speech tagging in Medieval Romance languages, comparing them against traditional taggers. The study demonstrates that LLM-based approaches with fine-tuning and cross-lingual transfer learning significantly outperform conventional methods, offering practical applications for digital humanities research on historical texts.

A Communication-Theoretic Framework for LLM Agents: Cost-Aware Adaptive Reliability

Researchers present a communication-theoretic framework that unifies LLM reliability techniques (retry, majority voting, self-consistency) under classical information theory, introducing a cost-aware router that achieves 56% lower costs than fixed approaches while maintaining quality. The work demonstrates that no single reliability technique dominates across all tasks, supporting dynamic per-task allocation strategies.

Field-Localized Forgery Detection for Digital Identity Documents

Researchers introduce FLiD, a lightweight deep learning framework that detects forged identity documents by analyzing specific fields like faces and text rather than entire documents. The method achieves superior accuracy to existing general-purpose forensics tools while using 13x fewer parameters, addressing a critical vulnerability in remote identity verification systems.

Formal Policy Enforcement for Real-World Agentic Systems

Investigating Anisotropy in Visual Grounding under Controlled Counterfactual Perturbations

Researchers investigate why visual grounding models fail when image captions are semantically mismatched, hypothesizing that embedding anisotropy may be responsible. Testing two transformer-based models with different embedding geometries reveals no meaningful correlation between cosine similarity and approximation errors, suggesting the problem requires investigation of deeper geometric properties.

Fairness of Explanations in Artificial Intelligence (AI): A Unifying Framework, Axioms, and Future Direction toward Responsible AI

Researchers present a unified framework addressing a critical gap between algorithmic fairness and explainable AI (XAI): models can produce fair outputs while employing biased reasoning processes. The study introduces the concept of 'procedural bias' and proposes a conditional invariance framework to formalize and audit explanation fairness, establishing the first comprehensive taxonomy and evaluation workflow for this emerging field.

Phase Transitions in Affective Meaning Divergence: The Hidden Drift Before the Break

Researchers formalize 'affective meaning divergence' (AMD)—the divergence in emotional interpretation of shared words between conversation partners—and demonstrate that it undergoes a critical phase transition before conversational breakdown. Using game-theoretic modeling and empirical analysis of 652 conversations, they show that AMD exhibits critical-slowing-down signatures predictive of relationship rupture, outperforming toxicity and sentiment baselines.

A Quantum Inspired Variational Kernel and Explainable AI Framework for Cross Region Solar and Wind Energy Forecasting

Researchers have developed a hybrid forecasting framework combining classical machine learning, quantum-inspired variational kernels, and generative AI to predict solar and wind energy generation across different geographic regions. The system achieves competitive performance with classical baselines while demonstrating superior ability to distinguish between calm and stormy weather patterns, with potential applications for power grid management and renewable energy optimization.

Towards Cross-lingual Values Judgment: A Consensus-Pluralism Perspective

A Geometric Perspective on Next-Token Prediction in Large Language Models: Three Emerging Phases

Researchers have developed a geometric framework for understanding how large language models process information across their layers, identifying three distinct phases in next-token prediction: Seeding Multiplexing, Hoisting Overriding, and Focal Convergence. The study reveals that model depth primarily increases capacity for candidate disambiguation rather than adding fundamentally new computational stages.

CAMAL: Improving Attention Alignment and Faithfulness with Segmentation Masks

Researchers introduce CAMAL, a method that leverages segmentation masks to improve attention alignment and faithfulness in vision models across deep learning and reinforcement learning paradigms. The approach achieves over 35% improvements in attention faithfulness while maintaining or improving generalization performance without additional inference costs.

Evolutionary Ensemble of Agents

Researchers introduce Evolutionary Ensemble (EvE), a decentralized framework that organizes coding agents into a self-evolving system for algorithmic discovery. By co-evolving two populations—functional code solvers and agent guidance states—EvE autonomously discovered novel mechanisms for In-Context Operator Networks, demonstrating that dynamic agent adaptation outperforms static optimization approaches.

The Reciprocity Gradient

Researchers introduce the reciprocity gradient, a novel machine learning method that addresses the influence attribution problem in multi-agent strategic interactions. The approach backpropagates reward signals through estimated opponent policies without requiring reward shaping, enabling agents to learn context-sensitive cooperation strategies that outperform sample-based baselines.

Can We Formally Verify Neural PDE Surrogates? SMT Compilation of Small Fourier Neural Operators

Researchers demonstrate that Fourier Neural Operators (FNOs) used for PDE simulation can be formally verified using SMT solvers by exploiting their piecewise-linear structure once weights are fixed. While exact encoding provides sound proofs and counterexamples on small models, scalability remains limited, revealing a fundamental tradeoff between formal verification rigor and practical applicability for production neural operators.

Latency Analysis and Optimization of Alpamayo 1 via Efficient Trajectory Generation

Researchers have optimized Alpamayo 1, a reasoning-based autonomous driving system, by redesigning it from multi-reasoning to single-reasoning architecture while accelerating diffusion-based action generation. The optimization achieves a 69.23% latency reduction while maintaining trajectory diversity and prediction quality, demonstrating that system-level efficiency improvements are critical for practical autonomous driving deployment.