Hacker News - Newest: "LLM" hnrss.org

Semantic retrieval vs. grep for LLM code queries: 2.4x faster, 5.6x fewer tokens

castnettechnology.com·4w·Hacker News

composo-ai/llm-judge-criteria-ensembling: Criteria injection + ensembling for LLM-as-judge accuracy on RewardBench 2 — 83.6% with one-sentence criteria and k=8 ensembling.

github.com·4w·Hacker News

Live and Let AI: Former CIA officer says human spies matter more in the LLM age

theregister.com·4w·Hacker News

Music-bench: An LLM benchmark for reading printed music

github.com·4w·Hacker News

JSON Zen – compact JSON schema for sanity checks and LLM context

json-zen.sleekcms.site·4w·Hacker News

Enhanced Mycelium of Thought (EMoT): A Bio-Inspired Hierarchical Reasoning Architecture with Strategic Dormancy and Mnemonic Encoding

arxiv.org·5w·Hacker News

dhanushk-offl/hallx: Lightweight, model-agnostic hallucination risk engine for LLM outputs

github.com·4w·Hacker News

Show HN: A Homeostatic Logic-Funnel to Prevent RLHF Overrides in LLM Personas

zenodo.org·4w·Hacker News

Add emotional awareness to any LLM in 3 lines of Python

resonance-layer.com·4w·Hacker News

elly99-AI/MarCognity-AI: A research framework for structured LLM evaluation, claim verification and reflective reasoning mechanisms.

github.com·7w·Hacker News, Hacker News, Hacker News

Lemonade by AMD: a fast and open source local LLM server using GPU and NPU

lemonade-server.ai·4w·Hacker News

Retry and Fallback: How Edgee Keeps Your LLM Requests Reliable

edgee.ai·5w·Hacker News

An LLM’s not going to tell you no

derekkedziora.com·15w·Hacker News

Show HN: An unstructured data workspace for data transformations with LLM

usefolio.ai·5w·Hacker News

Dhwanil25/Agentis: Most AI tools give you one model and one answer. Agentis gives you a team. Open-source multi-agent platform across 12 LLM providers — watch agents think, collaborate, and deliver in real time.

github.com·5w·Hacker News

zhihuiyuze/PDF-Prompt-Injection-Toolkit: A red team / blue team toolkit for testing and detecting prompt injection attacks hidden inside PDF documents. 一个用于测试和检测 PDF 文档中隐藏的提示词注入攻击的红蓝对抗工具包。

github.com·5w·Hacker News

Grove: Distributed LLM Training over AirDrop

github.com·5w·Hacker News

The Shape of Code » 70% of new software engineering papers on arXiv are LLM related

shape-of-code.com·6w·Hacker News, Hacker News

lightfeed/extractor: Using LLMs and AI browser automation to robustly extract web data

github.com·5w·Hacker News

michK/Autoresearch-for-Research: An LLM forms hypotheses, writes code, runs PDE simulations, and iterates — achieving 7–24× improvements across five physics domains on a single CPU VPS.

github.com·5w·Hacker News

Log in to enable infinite scrolling