Sitemap - 2023 - Interconnects
Interconnects year in review: 2023
Interviewing Tri Dao and Michael Poli of Together AI on the future of LLM architectures
State-space LLMs: Do we need Attention?
Big Tech's LLM evals are just marketing
OpenAI’s shakeup and opportunity for the rest of us: openness, brain drain, and new realities
Reckoning with the Shoggoth of AI
How the Foundation Model Transparency Index Distorts Transparency
RLHF lit. review #1 and missing pieces in RLHF
Undoing RLHF and the brittleness of safe LLMs
The AI research job market shit show (and my experience)
Open, general-purpose LLM companies might not be viable
Can robotics take off like GenAI? Moravec's paradox vs. scaling laws
DALL·E 3 and multimodality as moats, correcting bad moat takes
Challenges operationalizing responsible AI in open RLHF research
Midjourney vs. Ideogram, ML product companies, preventing AI winter, DALL·E 3 tease
In defense of the open LLM leaderboard
AI researchers' challenges: atomic analogies and strained institutions
Cruise's collisions and adapting to AI
LLM products: measurement and manipulation
"If it's not fully closed ML, it's open" - is it?
Llama 2 follow-up: too much RLHF, GPU sizing, technical details
Llama 2: an incredible open LLM
LLM agents follow-up: exploration, RLHF, and more
LLM agents and integration dead-ends
Disinformation with LLMs: the distribution risk
Tesla Autopilot's negligence and regulation
Different development paths of LLMs
Open-source LLMs' harmlessness gap
Evaluating and uncovering open LLMs
Unfortunately, OpenAI and Google have moats
Beyond human data: RLAIF needs a rebrand
The next chapter: Interconnects
Growing needs for accessing state-of-the-art reward models
Behind the curtain: what it feels like to work in AI right now (April 2023)
The implicit dynamics of optimizing costs vs. rewards vs. preferences
GPT4: The quiet parts and the state of ML
AGI Roundup: Re-visiting Go; transitioning from narrow to general; multimodality & GPT4
The RLHF battle lines are drawn
"AI alignment" and uncalibrated discourse on AI
Three seasons of RL: Metaphor, tool, and framework
Scaling laws for robotics & RL: Not quite yet
Pretraining quadrupeds: a case study in RL as an engineering tool