Researcher calls out vision RL papers as p-hacked rather than scalable results
gowthami_s · x · 2026-10-20
Researcher Gowthami S criticizes the quality of academic vision RL papers, saying she passed on sharing two because they were 'pretty much p-hacked rather than scalable results.' She adds she tried those ideas herself and they don't work at scale — the papers package them with carefully tuned hyperparameters and cherry-picked results — and asks whether a DL-era equivalent of p-hacking exists.
Related event: Researcher Calls Out p-Hacking in Vision RL Papers(2 posts)→
More from Research
- New MULTIFINGROUND Benchmark Shows LLMs Recognize Financial Concepts but Fail at Full Evidence Extraction — TheFinAI · 2026-10-20
- Video Geometry Corrector fixes 3D distortions in AI-generated videos, NeurIPS paper — kwangmoo_yi · 2026-10-20
- Ex-GitHub CEO Sid Sijbrandij open-sources 125,091 radiology images and multi-omics from his cancer journey — OgundipeOre · 2026-10-20
- PrismML's Ternary Bonsai 2 squeezes Qwen3.5 VL models to ternary weights, 9B at 2GB — lmoroney · 2026-10-20
- MIT Tech Review: "We were looking at it wrong" — the claim upending the AI world — ResultBackground2450 · 2026-10-20
- Prime Intellect unveils INTELLECT Policy Optimization, fixing IcePop's ratio masking — xeophon · 2026-10-20