lab Mistral AI News · 1h ago · releases

Mistral releases Agentic Search for accurate, efficient AI retrieval

from English

Mistral has released Agentic Search, a multi-step retrieval layer that enables AI systems to navigate, read, and verify information within complex documents. This tool is available through the Mistral Search Toolkit and Libraries in Studio and Vibe.

The system uses five tools (search, open, navigate, read, grep) to allow models to inspect and refine results iteratively rather than relying on one-shot RAG.
On FinanceBench, Agentic Search improved correctness from 26.7% to 86%, a 3x gain, while reducing p90 latency by up to 39.6%.
On OfficeQA Pro, the approach achieved a +45.6 point gain (6.3% to 51.9%) for GLM-5.2 and reduced token consumption by up to one-third.
The solution supports sensitive domain-specific data and works with existing indexes without requiring fine-tuning or model-specific training.

Agentic Search helps enterprises improve AI accuracy on dense, table-heavy documents like financial filings and legal contracts while reducing latency and token usage.

Importance 2/3 New harness with differentiators Trust 2/3 Mistral AI News Mistral AI AI agents Retrieval & RAG

Read original

arxiv arXiv cs.CL · 64d ago

MODE-RAG: Evaluating and Reducing Hallucinations in M-RAG

MODE-RAG proposes a multi-agent system using Variational Free Energy to dynamically gate interventions and reduce cross-modal hallucinations in retrieval-augmented generation. It integrates Monte Carlo Tree Search and logit perturbations to address causal fabrications and sycophancy, with dedicated agents ensuring factual verification and formatting stability. Evaluated via ModeVent, a subset of MultiVent, the system significantly improves robustness against logical fabrications.

media Latent Space · 2d ago · 1 view

Glean uses model routing to cut costs and select optimal LLMs

Driven by rising frontier model costs and the power of open-weight models, Glean employs model routing to dynamically select the most appropriate AI model for each enterprise task. This approach allows the platform to avoid using large language models for simple operations while optimizing performance and expense for complex queries.

media Hugging Face Forums · 3d ago · 4 views

AlphaAvatar v0.6.6 introduces event-driven memory, audio perception, and unified backend contracts

AlphaAvatar v0.6.6 shifts the framework's focus toward clearer runtime contracts and modular capabilities rather than introducing a single large feature. The release expands Environment Memory to include audio segments alongside visual observations and transitions memory updates from periodic polling to an event-driven model.

media Hugging Face Forums · 15d ago · 28 views

Aletheion releases ASM-CM, a compact memory architecture for persistent AI agents

Felipe Maya Muniz of AletheionAGI introduces ASM-CM (Aletheion Compact Memory Model), an experimental architecture that separates memory from language to allow agent history to grow without increasing active computational memory.

media Hugging Face Forums · 17d ago · 24 views

IntelShed combines hybrid RAG, GNN entity resolution, federated learning, and LLM multi-agent orchestration

The author presents IntelShed, an open-source system for open-source intelligence (OSINT) that aggregates 50 public data feeds to produce automated 24-hour security briefings. The architecture integrates hybrid retrieval, GNN-based entity resolution, federated learning with differential privacy, and LLM-compiled multi-agent orchestration.