media Hugging Face Forums · 1h ago · open_models

Researcher releases Ghost Search, an MCP-native Tor search layer for RAG

from English

A researcher has open-sourced Ghost Search, a privacy-preserving meta-search aggregator designed to reduce reliance on paid APIs for Retrieval-Augmented Generation workflows. The project combines 14 onion search engines with an MCP server interface, allowing direct integration into LLM agents without backend infrastructure or per-query fees.

Features include RAM-only operation with no database persistence, BM25 ranking with engine agreement boosts, and STIX 2.1 export capabilities.
The system uses NVIDIA Nemotron-3-Super-120B as a primary model with an Ollama fallback, enabling fully local execution on consumer hardware.
An Android app is currently in F-Droid review, offering a backend-free experience for mobile users.

The author is seeking research collaborators to help develop a search ranking benchmark dataset, design retrieval evaluation experiments, and conduct privacy audits of the Tor-based architecture.

Importance 1/3 New harness with differentiators Trust 1/3 Hugging Face Forums Evaluation & benchmarks Retrieval & RAG

Read original

media Hugging Face Forums · 19h ago

SplatRAG 3+ achieves SciFact R@10 0.8804 with slang in memory

SplatRAG 3+ demonstrates that a local Gaussian-splat memory can maintain high retrieval accuracy on the SciFact benchmark while simultaneously storing 5,000 Urban Dictionary slang rows without degradation.

arxiv arXiv cs.CL · 4d ago · 12 views

Nemotron-3-Embed closes commercial tax gap in MuSiQue retrieval benchmarks

A new audit of multi-hop retrieval benchmarks reveals that NVIDIA's Nemotron-3-Embed-8B has closed the performance gap between commercially licensed embedders and the non-commercial NV-Embed-v2 anchor on the MuSiQue benchmark. Released on July 16, 2026, this model is the only commercially licensed, free-to-self-host option that performs indistinguishably from the academic standard.

media Hugging Face Forums · 30d ago · 23 views

Author proposes open k registry for evidence saturation in local LLMs

The author has published a pilot paper and an open registry on Hugging Face to measure the evidence-saturation point (k*) of language models. This metric determines the optimal number of evidence fragments to inject into prompts, as adding more context does not monotonically improve correctness and can increase cost or epistemic contamination.

arxiv arXiv cs.CL · 37d ago · 11 views

LLM-as-a-judge signals fail to optimize closed-loop table recognition

A study using FinTabNet and OmniDocBench demonstrates that LLM-as-a-judge signals are insufficient for optimizing closed-loop regeneration in table recognition. The research reveals that judge scores frequently tied, rankings were not reproducible, and the system failed to recover better candidates produced by iteration.

arxiv arXiv cs.CL · 37d ago · 12 views

DeepStress stress-tests deep search agents against unreliable evidence

The authors introduce DeepStress, a framework for stress-testing deep search agents by replacing their retrieval modules with a controlled synthetic environment. This approach allows researchers to manipulate the frequency and quality of evidence across three dimensions: trustworthiness, relevance, and factuality.