Rethinking KV Cache Eviction via a Unified Information-Theoretic Objective
arXiv:2604.25975v1 Announce Type: new
Abstract: Key-value (KV) caching is essential for large language model inference, yet its memory overhead poses a critical bottleneck for long-context generation. Existing eviction policies predominantly rely on empirical heuristics, lacking a rigorous theoreti...
Operating-Layer Controls for Onchain Language-Model Agents Under Real Capital
arXiv:2604.26091v1 Announce Type: new
Abstract: We study reliability in autonomous language-model agents that translate user mandates into validated tool actions under real capital. The setting is DX Terminal Pro, a 21-day deployment in which 3,505 user-funded agents traded real ETH in a bounded on...
Distill-Belief: Closed-Loop Inverse Source Localization and Characterization in Physical Fields
arXiv:2604.26095v1 Announce Type: new
Abstract: {Closed-loop inverse source localization and characterization (ISLC) requires a mobile agent to select measurements that localize sources and infer latent field parameters under strict time constraints.} {The core challenge lies in the belief-space ob...
Evaluating Strategic Reasoning in Forecasting Agents
arXiv:2604.26106v1 Announce Type: new
Abstract: Forecasting benchmarks produce accuracy leaderboards but little insight into why some forecasters are more accurate than others. We introduce Bench to the Future 2 (BTF-2), 1,417 pastcasting questions with a frozen 15M-document research corpus in whic...
Hierarchical Multi-Persona Induction from User Behavioral Logs: Learning Evidence-Grounded and Truthful Personas
arXiv:2604.26120v1 Announce Type: new
Abstract: Behavioral logs provide rich signals for user modeling, but are noisy and interleaved across diverse intents. Recent work uses LLMs to generate interpretable natural-language personas from user logs, yet evaluation often emphasizes downstream utility,...
OMEGA: Optimizing Machine Learning by Evaluating Generated Algorithms
arXiv:2604.26211v1 Announce Type: new
Abstract: In order to automate AI research we introduce a full, end-to-end framework, OMEGA: Optimizing Machine learning by Evaluating Generated Algorithms, that starts at idea generation and ends with executable code. Our system combines structured meta-prompt...
Bootstrapping Sign Language Annotations with Sign Language Models
AI-driven sign language interpretation is limited by a lack of high-quality annotated data. New datasets including ASL STEM Wiki and FLEURS-ASL contain professional interpreters and 100s of hours of data but remain only partially annotated and thus underutilized, in part due to the prohibitive costs...
STARFlow-V: End-to-End Video Generative Modeling with Normalizing Flows
Normalizing flows (NFs) are end-to-end likelihood-based generative models for continuous data, and have recently regained attention with encouraging progress on image generation. Yet in the video generation domain, where spatiotemporal complexity and computational cost are substantially higher, stat...
Synthetic task scaling introduces a new training approach where AI agents learn through experience, closing the gap between knowledge and execution. Are you ready for the rise of the AI scientist?
GCA-BULF: A Bottom-Up Framework for Short-Term Load Forecasting Using Grouped Critical Appliances
arXiv:2604.24766v1 Announce Type: new
Abstract: With the rise of time-of-use and tiered electricity pricing, energy consumers are encouraged to adopt peak-shifting strategies by automatically controlling high-power appliances. These help lower energy costs while enhancing the power grid's stability...
Liquid Neural Network Models for Natural Gas Spot Price Time-Series Forecasting
arXiv:2604.24788v1 Announce Type: new
Abstract: Natural gas is undoubtedly an essential component of the global energy system. Accurate short-term forecasting of natural gas price is challenging due to pronounced volatility driven by seasonal demand patterns, geopolitical developments, and shifting...
Architecture Determines Observability in Transformers
arXiv:2604.24801v1 Announce Type: new
Abstract: Autoregressive transformers make confident errors, but activation monitoring can catch them only if the model preserves an internal signal that output confidence does not expose. This preservation is determined by architecture and training recipe. We ...
Co-Director: Agentic Generative Video Storytelling
arXiv:2604.24842v1 Announce Type: new
Abstract: While diffusion models generate high-fidelity video clips, transforming them into coherent storytelling engines remains challenging. Current agentic pipelines automate this via chained modules but suffer from semantic drift and cascading failures due ...
Latent Agents: A Post-Training Procedure for Internalized Multi-Agent Debate
arXiv:2604.24881v1 Announce Type: new
Abstract: Multi-agent debate has been shown to improve reasoning in large language models (LLMs). However, it is compute-intensive, requiring generation of long transcripts before answering questions. To address this inefficiency, we develop a framework that di...
S-SONDO: Self-Supervised Knowledge Distillation for General Audio Foundation Models
arXiv:2604.24933v1 Announce Type: new
Abstract: General audio foundation models have recently achieved remarkable progress, enabling strong performance across diverse tasks. However, state-of-the-art models remain extremely large, often with hundreds of millions of parameters, leading to high infer...
Adaptive Prompt Embedding Optimization for LLM Jailbreaking
arXiv:2604.24983v1 Announce Type: new
Abstract: Existing white-box jailbreak attacks against aligned LLMs typically append discrete adversarial suffixes to the user prompt, which visibly alters the prompt and operates in a combinatorial token space. Prior work has avoided directly optimizing the em...
Assessing Y-Axis Influence: Bias in Multimodal Language Models on Chart-to-Table Translation
arXiv:2604.24987v1 Announce Type: new
Abstract: Chart-to-table translation converts chart images into structured tabular data. Accurate translation is crucial for Multimodal Language Model (MLM) to answer complex queries. We observe imbalances in the number of images across different aspects of the...
Adaptive Thinking: Large Language Models Know When to Think in Latent Space
Recent advances in large language models (LLMs) test-time computing have introduced the capability to perform intermediate chain-of-thought (CoT) reasoning (thinking) before generating answers. While increasing the thinking budget yields smooth performance improvements at inference time, the relatio...
DSO: Direct Steering Optimization for Bias Mitigation
Generative models are often deployed to make decisions on behalf of users, such as vision-language models (VLMs) identifying which person in a room is a doctor to help visually impaired individuals. Yet, VLM decisions are influenced by the perceived demographic attributes of people in the input, whi...