Practitioners Community
385 resources · page 6 of 13
Context Rot: How Increasing Input Tokens Impacts LLM Performance (Chroma)
by Chroma
Chroma research report Context Rot: how increasing input length degrades LLM performance.
Jeremy Howard (@jeremyphoward)
Jeremy Howard's site: fast.ai co-founder, with links to courses and writing.
@jeremyphoward (Jeremy Howard)
X account of Jeremy Howard, co-founder of fast.ai and Answer.AI: posts on deep learning and AI tools.
@jerryjliu0 (Jerry Liu)
by Jerry Liu
X account of Jerry Liu, LlamaIndex co-founder and CEO: posts on document parsing, RAG and agents.
Jesse Vincent (@obra)
Jesse Vincent's blog, author of the superpowers skills framework for Claude Code.
Jim Fan (@DrJimFan)
by Jim Fan
NVIDIA GEAR lab, co-led by Jim Fan: generalist embodied agent research including the GR00T humanoid models.
@jobergum (Jo Kristian Bergum)
X account of Jo Kristian Bergum: posts on retrieval, ColBERT and RAG evaluation.
Joe Carlsmith (@jkcarlsmith)
Joe Carlsmith's site: essays on AI risk, alignment and philosophy; works on Claude's constitution at Anthropic.
JohannesGaessler GitHub profile
Johannes Gäßler's GitHub profile: contributor to llama.cpp and ggml CUDA performance work.
John Yang (@jyangballin)
by John Yang
SWE-agent site (John Yang and co-authors): agent for solving GitHub issues, alongside SWE-bench and SWE-smith.
Jonas Andrulis / Aleph Alpha
by Aleph Alpha
Aleph Alpha site: German company founded by Jonas Andrulis, building language models for enterprises and public institutions.
Jonas Geiping (@jonasgeiping)
Jonas Geiping's academic page: research on recurrent-depth latent reasoning models such as Huginn, ML security and privacy.
@julien_c (Julien Chaumond)
X account of Julien Chaumond, Hugging Face co-founder and CTO: posts on the Hub and open models.
Junyang Lin (@JustinLin610)
by Junyang Lin
Junyang Lin's page: researcher on language, multimodal and agentic models, formerly Qwen team lead, with papers and code.
Justine Tunney (@JustineTunney)
Mozilla llamafile GitHub repo, started by Justine Tunney: distribute and run LLMs as a single executable file.
@justinkan (Justin Kan)
by Justin Kan
X account of Justin Kan, Twitch co-founder: posts on startups and AI video projects.
@JustinLin610 (Junyang Lin)
by Junyang Lin
X account of Junyang Lin, former Qwen technical lead: posts on Qwen models and agent research.
@jxnlco (jason)
by jason
X account of Jason Liu (jxnl), OpenAI and Instructor library creator: posts on structured outputs and RAG.
@kaifulee (Kai-Fu Lee 李开复)
X account of Kai-Fu Lee, founder of 01.AI (Yi models): posts on AI in China and globally.
@rybolos_channel (Telegram)
Kali Novskaya's Telegram channel: Russian-language posts on neural networks and LLM development from an LLM team lead.
@karpathy (Andrej Karpathy)
X account of Andrej Karpathy: posts on LLMs, training and AI education such as nanoGPT and nanochat.
Kent C. Dodds (@kentcdodds)
Kent C. Dodds' Epic AI Pro: courses on building AI-powered applications and MCP.
Interconnected (Kevin Xu)
Kevin Xu's Interconnected: bilingual English–Chinese analysis of technology, investing and geopolitics, including AI.
Khoj
by Khoj AI
Khoj GitHub repo: self-hostable AI assistant that answers from the web and personal documents.
@kiwicopple (Paul Copplestone)
X account of Paul Copplestone, Supabase CEO: posts on Postgres, Supabase and AI app backends.
OpenPipe blog
by OpenPipe
OpenPipe blog archive (now part of Weights & Biases and CoreWeave): posts on agent RL training (ART).
Kyle Lo (@kylelostat)
by Kyle Lo
Kyle Lo's site: AI researcher on language model data, evaluation and data-centric methods, formerly at Ai2.
Lance Martin (@RLanceMartin)
by Lance Martin
Lance Martin's blog: posts on context engineering for agents and LangGraph tutorials.
LatamGPT (CENIA)
LatamGPT site: open language model built in Latin America for Latin American languages and contexts, led by Chile's CENIA.
@lateinteraction (Omar Khattab)
by Omar Khattab
X account of Omar Khattab, MIT professor and DSPy creator: posts on ColBERT, DSPy and GEPA.
Latent Space
by Latent Space
Latent Space: AI engineering newsletter and podcast on how labs build agents, models and infrastructure.
Leandro von Werra (@lvwerra)
Hugging Face TRL GitHub repo, led by Leandro von Werra: library for training language models with RL and preference methods.