TGL: AI Frontier Paper Club
Touch Grass Later is a recurring gathering for people who actually read the papers — or want to start.
Every session, we pick 1–2 papers from the AI frontier (think: reasoning models, scaling laws, agents, alignment, infra) and work through them together. No lectures, no slides. Just a room of people stress-testing ideas against each other.
Who it's for: Engineers, researchers, and technical founders who want to stay sharp on what's actually moving the field — not just the Twitter takes.
Format: One person leads a 15-min walkthrough of the paper. Then we go long on questions, implications, and what it means for what we're building.
The theme of this session will be on how long-horizon agents stay coherent. We're covering three papers that attack the same problem from different angles.
You don't need to have read them cover to cover. Skimming the abstract and intro of each is enough to get you into the discussion.
Paper 1 — EverMemOS (conceptual)
A self-organizing memory operating system for LLMs. The core idea: memory shouldn't be a flat retrieval store — it should have a lifecycle. Raw episodes get structured into semantic units, consolidated into thematic scenes, then reconstructively retrieved.
https://arxiv.org/abs/2601.02163
Paper 2 — GenericAgent A token-efficient self-evolving agent built around context information density maximization. The argument is that 30k tokens of high-density context beats 1M tokens of diluted context. Four mechanisms: minimal toolset, hierarchical on-demand memory, self-evolution via SOPs, and active context compression.https://arxiv.org/abs/2604.17091
Paper 3 — FluxMem Models memory as a heterogeneous graph that evolves its topology through feedback. The key move: while a task is running, it prunes bad connections and expands missing ones in real time.
https://arxiv.org/abs/2605.28773
Come prepared to read. Come ready to be wrong.