• your topics
• add yours

A LLM trained only on data from certain time periods to reduce modern bias

Show HN: EdgeVec – Sub-millisecond vector search in the browser (Rust/WASM)

Runmat

Testing a cheaper laminar flow hood

Show HN: Meal Tracker – Weekend hack using LLM to get food macros from photos

Show HN: Storyloom – Deterministic Storytelling Framework

Two Failures of Self-Consistency in the MultiStep Reasoning of LLMs (2024) [pdf]

A Glance at GPU Goodness in Java: LLM Inference with TornadoVM

Show HN: I'm building an open-source Amazon

Deciduous: A tool to work better LLMs using a decisions DAG and insight querying

Show HN: Help validate startup ideas with synthetic customer interviews

'Dead Cells' Free for a Limited Time on Epic Games' iOS Store in the EU

Show HN: CyberLoop – A control-loop architecture for stable LLM agents

Show HN: Quorum – Multi-agent CLI debates (AutoGen back end and React/Ink TUI)

Beyond Data Filtering: Knowledge Localization for Capability Removal in LLMs

Benchmark that evaluates LLMs using 759 NYT Connections puzzles

Vectorized MAXSCORE over WAND, especially for long LLM-generated queries