Lab Lands 9 Papers at NeurIPS, From Optimizers to Multi-Agent Self-Evolution
flosalim · x · 2026-09-25
A researcher (flosalim) announced that his group got 9 papers accepted at NeurIPS, promising a dedicated post for each (GlucoFM already covered). The list includes:
- FOGO: a forgetting-aware orthogonalization optimizer
- GlucoFM: a glucose-related foundation model work with the Google team
- MAGE: multi-agent self-evolution with co-evolutionary knowledge graphs
- Robust residual correction via selective deployment for time-series forecasting
- AnyMo: geometry-aware, setup-agnostic modeling of in-the-wild human motion
- Programmatic policies discovered from RL-based traffic signal controllers
- OOD detection in continual learning (collab with Melbourne's Erfani/Leckie)
Plus two evaluation-track papers: a retrieval-reranking decomposition for root-cause-analysis failure, and evaluating graded cognitive diversity in LLM persona simulation. The team heads to Sydney for the conference.
Related event: One team lands 9 papers at NeurIPS 2026(3 posts)→
More from Research
- leogao: alignment is spiky too — models may be aligned in some domains, misaligned in others — nabla_theta · 2026-09-25
- Fine-tuning π0.5 on a real manufacturing task: <1 hour of DAgger data drove big success-rate gains — DominiqueCAPaul · 2026-09-25
- MindGames, a benchmark for detecting lying AI agents, accepted at NeurIPS after 30K-game competition — LChoshen · 2026-09-25
- Alignment Is Likely Spiky Too: Models May Be Aligned in Some Domains, Misaligned in Others — nabla_theta · 2026-09-25
- After 8 years in AI formalization, Kaiyu Yang asks what happens when it gets cheap — KaiyuYang4 · 2026-09-25
- GRAM accepted to NeurIPS 2026: recursive reasoning becomes stochastic latent trajectories — SungjinAhn_ · 2026-09-25