AI for science: KDD 2026 vs NeurIPS 2024–2025 vs AI4Science workshops

Compares

KDD 2026 AI for Sciences track as “KDD 2026 AI for Sciences track” · run 1 · openai/gpt-oss-120b
NeurIPS 2024–2025 main track, ML-for-science areas as “NeurIPS 2024–2025 main track, ML-for-science areas” · run 1 · openai/gpt-oss-120b
AI for Science workshops 2021–2025 as “AI for Science workshops 2021–2025” · run 1 · openai/gpt-oss-120b

Not a clean comparison: The analyses differ in upstream fork version (fixes the extraction prompt) (KDD 2026 AI for Sciences track: ad1bd3df2b45417c2e1782b752212972bec2ed11; NeurIPS 2024–2025 main track, ML-for-science areas: c8c35f989996b40f13dcb16d1b4ceb46bdb5243f; AI for Science workshops 2021–2025: c8c35f989996b40f13dcb16d1b4ceb46bdb5243f). Claims extracted differently are not directly comparable.

Open Compare in the inspector and pick the cohorts under “Conference”. Everyone who opens this comparison needs view rights on every compared analysis; abstracts are not included.

Method and citation

Method

This analysis runs the original Drift Inspector methodology: Atomic Contribution Claims extracted from titles and abstracts by an LLM, SPECTER2 embeddings, UMAP and HDBSCAN clustering, and per-year document frequency.

Vsevolod Karimov, Stepan Ostarkov, Anastasia Poroshina, Anatoly Frolov, Alexander Panchenko. Drift Inspector: Exploring and Measuring Scientific Drift with Atomic Contribution Claims. 2026. arXiv:2609.39710 · code (MIT)