SIGNAL
Tracking the global AI frontier — labs · research · agents · policy
Frontier Signal

Research

Curated daily papers, lab research blogs and national AI research programs across the US, China, EU, UK, Japan and beyond.

Curated daily papers, lab research blogs and national AI research programs across the US, China, EU, UK, Japan and beyond.

Latest in Research

30 stories
Accelerating Disaggregated RL for Visual Generative LLMs with Diffusion-Based Parallelism and Trainer-Assisted Generation
Research

Accelerating Disaggregated RL for Visual Generative LLMs with Diffusion-Based Parallelism and Trainer-Assisted Generation

Reinforcement learning (RL) has become a dominant post-training paradigm, driving the emergence of high-performance RL systems such as veRL for autoregressive l…

Structural Kolmogorov-Arnold Convolutions: Learnable Function on the Values or the Filter Shape as Parameter-Efficient Alternative to Per-Edge Convolutional KANs
Research

Structural Kolmogorov-Arnold Convolutions: Learnable Function on the Values or the Filter Shape as Parameter-Efficient Alternative to Per-Edge Convolutional KANs

Convolutional Kolmogorov--Arnold Networks (KANs) replace the fixed weights of a convolutional kernel with learnable univariate functions. The dominant formulati…

ComputeFHE: A Privacy-Preserving General-Purpose Computation Library
Research

ComputeFHE: A Privacy-Preserving General-Purpose Computation Library

Fully Homomorphic Encryption (FHE) enables computations to be performed directly on encrypted data while preserving data confidentiality. However, its practical…

Entity Resolution via Batched Oracle Queries
Research

Entity Resolution via Batched Oracle Queries

We consider an oracle that processes a limited batch of records at a time and clusters those that refer to the same real-world entity. We study how to interroga…

Cycle-Consistent Neural Explanation of Formal Verification Certificates
Research

Cycle-Consistent Neural Explanation of Formal Verification Certificates

Formal verification produces machine-checkable certificates that attest to the satisfaction or violation of temporal properties, yet these certificates remain o…

Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning
Research

Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning

Experience-driven self-evolution is critical for large language model (LLM) agents to improve through open-world interaction. However, existing experience learn…

Detecting AI Coding Agents in Open Source: A Validated Multi-Method Census of 180 Million Repositories
Research

Detecting AI Coding Agents in Open Source: A Validated Multi-Method Census of 180 Million Repositories

Generative AI coding agents are entering the open-source supply chain, yet their diverse and often invisible traces leave their prevalence poorly understood. We…

MedPCFM: Improving Medical Point Cloud Completion by Integrating Point Transformers and Flow Matching
Research

MedPCFM: Improving Medical Point Cloud Completion by Integrating Point Transformers and Flow Matching

Medical point cloud completion is important for anatomical reconstruction and downstream clinical workflows, yet generative modeling in this setting remains ins…

ReM-MoA: Reasoning Memory Sustains Mixture-of-Agents Scaling
Research

ReM-MoA: Reasoning Memory Sustains Mixture-of-Agents Scaling

Mixture-of-Agents (MoA) architectures improve inference-time scaling by organizing multiple LLM agents into layered reasoning pipelines. However, existing MoA v…

S1-Omni-Image: A Unified Model for Scientific Image Understanding, Generation, and Editing
Research

S1-Omni-Image: A Unified Model for Scientific Image Understanding, Generation, and Editing

We present S1-Omni-Image, an open-weight unified multimodal model for scientific image understanding, generation, and editing. Unlike general-purpose image gene…

CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference
Research

CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference

Long-context large language model (LLM) inference is increasingly constrained by the memory footprint and decoding cost of key-value (KV) caches, limiting susta…

Advancing WordArt-Oriented Scene Text Recognition: Datasets and Methods
Research

Advancing WordArt-Oriented Scene Text Recognition: Datasets and Methods

WordArt (artistic text) features highly customized fonts, textures, and layouts, making WordArt-oriented scene TExt Recognition (WATER) substantially more chall…

Red-Teaming the Agentic Red-Team
Research

Red-Teaming the Agentic Red-Team

The use of agentic systems to perform offensive security operations has moved from a theoretical possibility to a commoditized capability. However, while the co…

AGORA: An Archive-Grounded Benchmark for Agentic Workplace Document Reasoning
Research

AGORA: An Archive-Grounded Benchmark for Agentic Workplace Document Reasoning

Large language models are increasingly deployed as agents that reason over documents rather than answer from parametric knowledge. We study archive-grounded rea…

PointVG-R: Internalizing Geometric Reasoning in MLLMs for Precise Pointing Localization via Visual Chain of Thought
Research

PointVG-R: Internalizing Geometric Reasoning in MLLMs for Precise Pointing Localization via Visual Chain of Thought

Pointing-based visual grounding requires models to precisely locate target objects by deciphering complex spatial relationships between the visual scene and poi…

To Compare, or Not to Compare: On Methodological Practices in Evaluating Social Bias
Research

To Compare, or Not to Compare: On Methodological Practices in Evaluating Social Bias

As Large Language Models are increasingly deployed in critical applications, robustly evaluating their social biases is paramount. However, the current literatu…

Uncertainty-Aware Longitudinal Forecasting of Alzheimer's Disease Progression Using Deep Learning
Research

Uncertainty-Aware Longitudinal Forecasting of Alzheimer's Disease Progression Using Deep Learning

Longitudinal modelling of Alzheimer's disease progression is clinically useful only if it can describe not just the most likely next diagnosis, but how a patien…

Infinitesimal Causality
Research

Infinitesimal Causality

This paper introduces a categorical account of infinitesimal causality in Frobenius Markov categories equipped with tangent-bundle semantics. IDC captures the i…

Themis: An explainable AI-enabled framework for Reinforcement Learning with Human Feedback
Research

Themis: An explainable AI-enabled framework for Reinforcement Learning with Human Feedback

Training safe Reinforcement Learning (RL) systems is inherently challenging, with no guarantee of avoiding unwanted behaviors. The most effective defenses again…

The Warrant Gap: Claim-Conditioned Re-scoring for Fact-Checking
Research

The Warrant Gap: Claim-Conditioned Re-scoring for Fact-Checking

Fact-checking systems built on LLMs achieve high verdict accuracy on standard benchmarks, yet routinely output Supports labels whose cited evidence does not lic…

Extended pseudo-spectral physics-informed neural networks for phase-field models
Research

Extended pseudo-spectral physics-informed neural networks for phase-field models

Phase-field models play a central role in the continuum description of phase separation, in which the bulk free-energy density and the interfacial thickness par…

CN-NewsTTS Bench: a target-level automatic benchmark for raw-input Chinese news TTS pronunciation
Research

CN-NewsTTS Bench: a target-level automatic benchmark for raw-input Chinese news TTS pronunciation

Chinese news text contains dense written forms such as scores, hyphenated model names, ranges, unit symbols, percentages, English abbreviations, and mixed Chine…

UniDrive: A Unified Vision-Language and Grounding Framework for Interpretable Risk Understanding in Autonomous Driving
Research

UniDrive: A Unified Vision-Language and Grounding Framework for Interpretable Risk Understanding in Autonomous Driving

Recent multimodal large language models (MLLMs) have shown strong potential for autonomous driving scene understanding, yet existing methods still face a fundam…

Revealing Training Data Exposure in Vision Language Large Models via Parameter Gradients
Research

Revealing Training Data Exposure in Vision Language Large Models via Parameter Gradients

Vision-Language Large Models (VLLMs) trained on massive crawled corpora raise pressing copyright and data-provenance concerns. These concerns are particularly a…

Grad Detect: Gradient-Based Hallucination Detection in LLMs
Research

Grad Detect: Gradient-Based Hallucination Detection in LLMs

Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks, yet they remain prone to generating hallucinations. Detecting these…

Less is More: Quality-Aware Training Data Selection for Scientific Summarization
Research

Less is More: Quality-Aware Training Data Selection for Scientific Summarization

Scientific long-document summarization datasets commonly treat author-written abstracts as gold reference summaries, although their quality and alignment with t…

DiffusionBench: On Holistic Evaluation of Diffusion Transformers
Research

DiffusionBench: On Holistic Evaluation of Diffusion Transformers

Diffusion transformer (DiT) research on image generation has converged to a single evaluation setup: class-conditional generation on ImageNet. While methods imp…

GroundEval: A Deterministic Replacement for LLM-as-Judge in Stateful Agent Evaluation
Research

GroundEval: A Deterministic Replacement for LLM-as-Judge in Stateful Agent Evaluation

Before letting an agent operate over real context, can you prove it used the right evidence? GroundEval turns that question into a deterministic test of what th…

Error Highways: Scaling Predictive Coding to Very Deep Networks
Research

Error Highways: Scaling Predictive Coding to Very Deep Networks

Predictive coding networks (PCNs) offer a biologically-plausible, local-learning alternative to back-propagation of errors (backprop). Nevertheless, they have r…

Learning Moral Diversity: Modelling Individual Perspectives in Moral Classification of Texts
Research

Learning Moral Diversity: Modelling Individual Perspectives in Moral Classification of Texts

Understanding moral values in social media text offers insight into moral judgement formation, and supervised NLP models trained on crowdsourced data have achie…