SignalIA
ENFR
← Archives

June 2026

2731 articles

GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> google /</span> skills

Google releases Skills, a collection of tools and capabilities for AI agents compatible with Google products and technologies. The project provides integrations and extended functionalities for multi-agent systems.

AI AgentsMulti-agentDeepMind
SIG
65
HYP
25
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> phuryn /</span> pm-skills

PM Skills Marketplace offers 100+ agentic skills, commands, and plugins spanning discovery, strategy, execution, launch, and growth phases.

AI AgentsToolsOpen source
SIG
35
HYP
55
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> biomejs /</span> biome

Biome is a toolchain for web projects providing formatting and linting capabilities via CLI and LSP. Open-source project aimed at code quality maintenance.

ToolsOpen sourceCode generation
SIG
45
HYP
15
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> idootop /</span> open-xiaoai

Open-source project enabling advanced voice listening capabilities for Xiaoai Speaker. Unlocks unlimited voice features on Xiaomi's smart speaker.

Open sourceVoice
SIG
35
HYP
45
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> 777genius /</span> agent-teams-ai

Multi-agent framework where you supervise autonomous AI teams via kanban board. Integrates 200+ models and 75+ LLM providers (Claude, Codex, OpenCode). Agents collaborate, review each other's work, and execute tasks independently.

AI AgentsMulti-agentClaude
SIG
45
HYP
65
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> xerrors /</span> Yuxi

Multi-tenant Agent Harness platform integrating LightRAG knowledge base and knowledge graphs. Built with LangChain + Vue + FastAPI, supports DeepAgents, MinerU PDF, Neo4j, MCP.

AI AgentsRAGMCP
SIG
65
HYP
25
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> langchain-ai /</span> deepagents

DeepAgents is a batteries-included agent framework from LangChain. It provides a complete architecture for building and deploying autonomous agents with native tool integration and workflows.

AI AgentsOpen sourceTools
SIG
45
HYP
35
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> google /</span> skills

Google releases Skills, a collection of tools and resources for building AI agents compatible with Google products and technologies. The project provides reusable capabilities and integrations for constructing multi-agent systems.

AI AgentsMulti-agentDeepMind
SIG
65
HYP
25
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> magenta /</span> magenta-realtime

Magenta RealTime 2 is Google's open-weights model for real-time live music generation. It enables low-latency music creation accessible as open-source.

DeepMindOpen source
SIG
75
HYP
25
GitHub Trending·

<svg aria-hidden="true" data-component="Octicon" height="16" viewBox="0 0 16 16" version="1.1" width="16" data-view-component="true" class="octicon octicon-repo mr-1 tmp-mr-1 color-fg-muted"> <path d="M2 2.5A2.5 2.5 0 0 1 4.5 0h8.75a.75.75 0 0 1 .75.75v12.5a.75.75 0 0 1-.75.75h-2.5a.75.75 0 0 1 0-1.5h1.75v-2h-8a1 1 0 0 0-.714 1.7.75.75 0 1 1-1.072 1.05A2.495 2.495 0 0 1 2 11.5Zm10.5-1h-8a1 1 0 0 0-1 1v6.708A2.486 2.486 0 0 1 4.5 9h8ZM5 12.25a.25.25 0 0 1 .25-.25h3.5a.25.25 0 0 1 .25.25v3.25a.25.25 0 0 1-.4.2l-1.45-1.087a.249.249 0 0 0-.3 0L5.4 15.7a.25.25 0 0 1-.4-.2Z"></path> </svg> <span data-view-component="true" class="text-normal"> Andyyyy64 /</span> whichllm

CLI tool to find the best-performing local LLM for your hardware. Ranked by real, recency-aware benchmarks rather than parameter count. One-command instant execution.

Open sourceToolsBenchmarks
SIG
65
HYP
35
arXiv cs.AI·

Hierarchical Semantic-Constrained Heterogeneous Graph for Audio-Visual Event Localization

HSCHG method for open-vocabulary audio-visual event localization. Constructs hierarchical heterogeneous graph in Euclidean space with segment and video-level nodes, applies selective cross-modal fusion with dual-threshold filtering, and projects representations into hyperbolic space with hierarchical entailment regularization. Outperforms existing methods on OV-AVEL benchmark.

VisionVoiceBenchmarks
SIG
72
HYP
15
arXiv cs.AI·

OpenSkill: Open-World Self-Evolution for LLM Agents

OpenSkill introduces a framework for open-world self-evolution of LLM agents without target-task supervision. The agent acquires grounded knowledge and verification signals from documentation and web resources, synthesizes transferable skills, and refines them through self-built virtual tasks. Results across 3 benchmarks show best automated pass rates with cross-model skill transfer.

AI AgentsReinforcement learningPapers
SIG
72
HYP
28
arXiv cs.AI·

Accelerated Fourier SAT (AFSAT): Fully Realising a GPU-based Symmetric Pseudo-Boolean SAT Solver

AFSAT is a GPU-accelerated solver for pseudo-Boolean satisfiability using continuous local search. Built with JAX, it leverages pure function composition, automatic vectorization, and JIT compilation for massive parallelization across candidate assignment batches. Improves numerical stability, runtime performance, and memory efficiency over the FastFourierSAT proof-of-concept.

BenchmarksInfrastructurePapers
SIG
72
HYP
15
arXiv cs.CL·

MADE: Beyond Scoring via a Multilingual Agentic Diagnosing Engine for Fine-Grained Evaluation Insights

MADE is a multilingual agentic diagnosing engine that decomposes post-evaluation analysis into planning, aggregate analysis, instance-level inspection, and grounded report synthesis. Tested on 33 model families, 11 benchmarks, and 26 languages (8.66M evaluation records), MADE outperforms strongest baselines by 47% in diagnosis quality and is preferred by human experts in 87.9% of comparisons.

AI AgentsMulti-agentEvals
SIG
78
HYP
25
arXiv cs.CL·

Tree-of-Experience: A Structured Experience-Management Solution for Self-Evolving Agents under Low-Repetition and Implicit-Reward Environments

New paper on structured experience management for LLM agents in implicit-reward environments. Introduces FinEvolveBench, a financial sentiment prediction benchmark with delayed, noisy feedback, and Tree-of-Experience (ToE), an experience organization and validation method that outperforms no-experience baselines.

AI AgentsReinforcement learningBenchmarks
SIG
72
HYP
28
arXiv cs.CL·

CRAFT: A Unified Counterfactual Reasoning Framework for Tabular Question Answering and Fact Verification

CRAFT is a unified counterfactual reasoning framework for tabular question answering and fact verification. The method reformulates these tasks as bidirectional verification processes, explicitly constructing declarative statements and their counterfactual variants. Results: improvements on WikiTQ and TabFact, reduced performance gaps across LLMs.

ReasoningBenchmarksPapers
SIG
72
HYP
18
arXiv cs.CL·

TA-RAG: Tone-Aware Retrieval-Augmented Generation for Peer-Support Health Communication

TA-RAG is a lightweight prompt-based RAG framework embedding tone control into RAG pipelines without fine-tuning. It operationalizes tone through four components: stigma-free rewriting, readability adjustment, recipient adaptation, and empathy rephrasing. Evaluated on HIV peer-support data (HOLA, NAPWHA), it improves communication quality while preserving content.

RAGPrompt engineeringAI safety
SIG
72
HYP
18
arXiv cs.CL·

Explain Like I'm 5 or Whatever I Choose: Evaluating the Interactive Potential of Language Model Responses

Evaluation study of LLMs (GPT-5.1, GPT-5 mini, Claude Sonnet 4.5 + Thinking, DeepSeek-V3.1) on their ability to generate multiple responses to the same scientific query while varying language complexity. On 98 queries, Claude Sonnet 4.5 maintains consistent complexity only 46% of the time. Evaluation framework based on formative study with 16 participants.

EvalsClaudeGPT
SIG
72
HYP
25
arXiv cs.CL·

Improving Cross-Lingual Factual Recall via Consistency-Driven Reinforcement Learning

PolyFact, a 100K multilingual factual QA dataset grounded in Wikidata across 12 languages, evaluates three approaches to improve cross-lingual factual consistency in Qwen-2.5-7B and OLMo-2-1124-7B. GRPO outperforms supervised fine-tuning by reducing language specialization in MLP layers and attention heads, promoting shared cross-lingual representations.

BenchmarksReinforcement learningQwen
SIG
82
HYP
18
arXiv cs.LG·

FAIR-Calib: Frontier-Aware Instability-Reweighted Calibration for Post-Training Quantization of Diffusion Large Language Models

FAIR-Calib introduces a post-training quantization (PTQ) method for diffusion large language models (dLLMs). The two-stage framework protects fragile frontier decisions by reweighting unstable hidden states, avoiding expensive diffusion rollouts. Results on LLaDA and Dream (W4A4) show reduced quantization errors and decision flips.

LlamaFine-tuningPapers
SIG
72
HYP
15
arXiv cs.AI·

Declarative Skills for AI Agents in Knowledge-Grounded Tool-Use Workflows

Comparative study of three orchestration paradigms for tool-using AI agents in customer-service workflows: declarative agents (natural-language skill files), imperative agents (state machines), and unscaffolded baseline. Tested on 5 LLMs and 2 retrieval regimes. Finding: retrieval quality is the dominant bottleneck; declarative skills consistently improve accuracy on procedural tasks when retrieval is high-quality.

AI AgentsRAGPrompt engineering
SIG
72
HYP
18
arXiv cs.CL·

The Dark Regulome: Disentangling Predictability from Regulation in Genomic Foundation Models

Study of genomic foundation models to identify regulatory elements in the dark genome (dark regulome) of gliomas. Authors separate sequence predictability from actual regulation via residualization-permutation across three models (Caduceus-Ph, HyenaDNA, Enformer). Results: 10kb proximal regulatory horizon, 3.3× enrichment in brain eQTLs, but LM-derived hierarchy non-reproducible.

BenchmarksPapersReasoning
SIG
72
HYP
15
arXiv cs.CL·

Progress-SQL: Improving Reinforcement Learning for Text-to-SQL via Progressive Rewards

Progress-SQL introduces a multi-turn reinforcement learning framework with progressive rewards for Text-to-SQL generation. The method proposes an Oracle-guided Diagnostic Tree (ODT) that abstracts SQL queries at clause level and provides progressive rewards measuring improvement from initial to final SQL. Evaluated on BIRD, Spider, and robustness variants.

Reinforcement learningCode generationReasoning
SIG
72
HYP
18
arXiv cs.CL·

When Better Codebooks Are Not Enough: Predictive Performance and Behavioral Reliability in LLM Political Event Coding

Study on political event coding with LLMs: higher accuracy does not ensure behavioral reliability. Expert codebooks optimized with clearer definitions, examples, and context improve performance, but models fail reliability tests under controlled variations (label names, codebook order, mappings). Accuracy alone is insufficient for social-science applications.

EvalsReasoningPrompt engineering
SIG
72
HYP
15