[ICML 2026] ByteDance's All-in-One Video Generation Model for Human-Object Interaction Video Generation

[ICML 2026] World-R1: Reinforcing 3D Constraints for Text-to-Video Generation

[ICML 2026] Official implementation for "DyPE: Dynamic Position Extrapolation for Ultra High Resolution Diffusion".

[ICML 2026 & ICLR 2026 AIWILD] Official Implementation of the CKA-Agent, "The Trojan Knowledge: Bypassing Commercial LLM Guardrails via Harmless Prompt Weaving and Adaptive Tree Search".

[ICML 2026] Orienting Latent Actions for Video World Modeling

[𝗜𝗖𝗠𝗟 𝟮𝟬𝟮𝟲] Dispersion loss counteracts embedding condensation and improves generalization in small language models

Search-Then-Verify natural language querying over time series databases (ICML 2026). Includes NLQTSBench.

[ICML 2026 Spotlight] Official PyTorch implementation of “SplAttN: Bridging 2D and 3D with Gaussian Soft Splatting and Attention for Point Cloud Completion”.

[ICML 2026] Scaling Continual Learning to 300+ Tasks with Bi-Level Routing Mixture-of-Experts

(ICML 2026) IDLM: Inverse-distilled Diffusion Language Models

Dreaming in Code for Curriculum Learning in Open-Ended Worlds (Accepted at ICML 2026)

[ICML 2026] TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting

Robust contextual dueling bandits with post-serving context, delayed feedback, and adversarial corruption (RLHF / preference learning) — ICML 2026

Code for "Jailbreaking Vision-Language Models Through the Visual Modality" (ICML 2026)

[ICML 2026 Spotlight] HELIX: Hybrid Encoding with Learnable Identity and Cross-dimensional Synthesis for Time Series Imputation

[ICML2026] Towards Feedback-to-Plan Decisions for Self-Evolving LLM Agents in CUDA Kernel Generation

[ICML 2026] VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer Authority

Protein Fold Classification at Scale: Benchmarking and Pretraining (ICML 2026 oral)

Official implementation of the ICML 2026 paper "DiLA: Disentangled Latent Action World Models".

Code for ICML 2026 paper experiments studying how deep sequence models tend to memorize through a learned geometric structure.