101k papers mined for trends: RLVR jumps to 110 ICLR papers, ~25% of tested tweaks don't help
latenticlabs · reddit · 2026-11-10
The latenticlabs team expanded their topic pages built from 10 years of ICML/NeurIPS/ICLR/CVPR/AAAI papers, now covering 101k full texts (1M 'tried X, got Y' findings) and most of ICML/CVPR/AAAI/ICLR 2026. Their year-over-year 'newspapers' reveal: LLM agents went from 1.2% to 4.4% of ICML; multimodal LLMs doubled at CVPR (5.4%→10.8%); RL with verifiable rewards went from zero to 110 ICLR papers; Qwen is the most-used base model across all three; and a quarter of tested changes didn't help by the papers' own numbers. Their Gaussian splatting example shows baselines are a mess — plain 3DGS PSNR on Mip-NeRF 360 is reported anywhere from 16 to 27.7, so 'beats 3DGS by 2 dB' means very different things. Data is AI-extracted and unverified; use as a starting point, not a leaderboard.
More from Research
- Claude Fable 5.2 tops VIALS-AA science visual reasoning benchmark at 67% with heavy tool use — ArtificialAnlys · 2026-11-10
- WVU researcher presents 10 years of tractometry work on Alzheimer's white matter vulnerability — PTenigma · 2026-11-10
- New cryo-EM structure: pSTAT1 bound to importin alpha 5 (PDB 11GJ) — chaitjo · 2026-11-10
- ARLoop: fully-open 1.2B looped transformer beats Huginn-3.5B and Qwen3.5-2B on all 8 benchmarks — hhsun1 · 2026-11-10
- RAGCap-Bench: A Fine-Grained Benchmark for Intermediate Capabilities in Agentic RAG — kalyan_kpl · 2026-11-10
- RBA gives LLaVA-OneVision-2 a 3x lossless training speed-up, merged upstream — liuziwei7 · 2026-11-10