Topic
AI Agents & Reasoning
Getting LLMs to plan, use tools, remember, and reason — what the agent papers actually demonstrate and how the patterns hold up in production.
Toolformer: What the Paper Actually Says
A 6.7B model beats GPT-3 175B on math by learning to use a calculator. Toolformer's self-supervised training pipeline is the interesting part — and it's more constrained than the demos imply.
Generative Agents: What the Paper Actually Says
The Generative Agents paper isn't just a simulation demo. It's a formal architecture for agent memory: stream, retrieval, reflection, and planning. Every agent framework you're using borrowed from it.
The AI agent observability stack — what to measure, what to use
Most teams reach for one observability tool and call it solved. Different tools measure different things — traces, hallucinations, cost, drift. Here's a map of the space and where to start.
Tree of Thoughts: What the Paper Actually Says About LLM Search
Chain-of-thought is one path through a reasoning problem. Tree of Thoughts is a search over many paths. The NeurIPS 2023 paper shows 74% vs 4% on certain benchmarks. The production story is harder.
ReAct: What the Reasoning + Acting Paper Actually Says
Every agent framework you've used is built on ReAct's thought/action/observation loop. The paper is more specific — and more fragile — than the abstractions suggest.
DeepSeek-R1: What Reinforcement Learning with Verifiable Rewards Actually Does
DeepSeek-R1 showed that a model can learn to reason—backtrack, self-correct, think for longer on harder problems—with no human-labeled reasoning traces. Just a reward signal that says 'right or wrong.' Here's what that means in practice.
Chain-of-Thought Prompting: What the Wei et al. Paper Actually Says
CoT prompting is everywhere. Most engineers treat it as a free latency-for-accuracy trade. The Wei et al. paper is more specific — and more limiting — than that.