github LlamaIndex · 2d ago · agents

Llama Index v0.14.25 resolves security alerts and adds async support

from English

Llama Index released version 0.14.25 on September 21, 2026, primarily addressing widespread security vulnerabilities across its ecosystem.

Resolved numerous security alerts in core and extension packages via pull request #22855.
Removed deprecated ipex-llm and optimum-intel IPEX integrations from the core module.
Added native async support to StructuredLLMRerank.
Fixed issues with metadata replacement fallback, streaming tool calls, and function tool retries.
Re-locked 93 manifests to clear stuck Dependabot alerts in embedding packages.

This update improves security posture for users by patching dependencies and modernizes the core library by removing obsolete integrations.

Importance 1/3 Trust 2/3 LlamaIndex Retrieval & RAG

Read original

media Hugging Face Forums · 1d ago · 10 views

YUCLAW 8.0.0 adds explicit snapshot mode for reproducible evidence demos

YUCLAW 8.0.0 introduces an explicit snapshot mode to ensure that command-line interface demonstrations produce byte-exact, reproducible output regardless of the host environment's network connectivity.

media Hugging Face Forums · 1d ago · 10 views

jbsalles experiments with SelMem, a selective memory system for LLMs

The author has developed a Rust project called SelMem to experiment with imperfect, path-dependent memory for large language models. This approach mimics human memory by allowing agents to compress, forget, and reconstruct memories over time rather than storing perfect copies of interactions.

lab OpenAI News · 2d ago · 18 views

V7 uses GPT-6 Astra to give AI agents institutional memory

V7 has integrated OpenAI's GPT-6 Astra into its V7 Go platform to provide AI agents with persistent "institutional memory" via a Context Graph, significantly improving retrieval accuracy on complex enterprise queries.

media Hugging Face Forums · 4d ago · 9 views

Semantic routing reduces Qwen2.5 input tokens by 41% and latency by 45% in 14K context stress test

A GPU stress test using frozen 4-bit Qwen2.5-7B-Instruct on HotpotQA bridge questions demonstrates that task-aware semantic routing significantly improves efficiency and accuracy under high-distractor conditions. The study compared full capped context against a query-aware semantic selector retaining 60% of the budget, revealing substantial gains in speed and precision.

media Hugging Face Forums · 5d ago · 12 views

jbsalles shares SelMem, an experimental Rust project for LLM selective memory

The author introduces SelMem, an early-stage experimental project written in Rust that explores whether LLM agents can develop persistent behavioral divergence through a reconstructive memory system. This system allows agents to forget, distort, and consolidate experiences over time, potentially leading to different internal narratives even with identical histories.