How do you attribute agent gains: model training or infrastructure?
jrxna · reddit · 2026-10-16
In the Oct 12 Cheeky Pint episode, Zuckerberg and Alexandr Wang describe agent behavior as shaped through model training while also discussing surrounding privacy and security infrastructure. The poster asks how to evaluate which layer drives better results: holding tools and tasks fixed while swapping models is useful, but models may have been trained around different tool interfaces. They propose running both a shared setup and each model's preferred setup, reporting task completion, permission violations, and recovery from failed actions separately.
More from coding & agent
- 'LLM code is sloppy' rebutted with real ffmpeg code: identical if/else branches in a top library — kingslyj · 2026-10-16
- Agent actions need checks outside the model: NVIDIA demos lean on guardrails amid wave of real agent incidents — Creamy-And-Crowded · 2026-10-16
- Vitalik builds "paranoid mode" agent wrapper with Tor, zkAPI and local prompt rewriting — DavideCrapis · 2026-10-16
- Agent fine-tunes Qwen3.5-0.8B with synthetic data: 4.7%→76.3% accuracy, 11.9× faster — ben_burtenshaw · 2026-10-16
- Beam-501B agent demo: solo beats 5-agent team on ProgramBench rebuild task — ben_burtenshaw · 2026-10-16
- Beam agent builds BeamSpeak, a local Parakeet dictation app for macOS — ben_burtenshaw · 2026-10-16