Geometric prompting framework for remote sensing open-vocabulary semantic segmentation.

Open-source genetic programming system for interpretable symbolic regression on small, wide datasets.

Open-source multi-agent pipeline for automated systematic review and meta-analysis.

Open-source LLM-agent pipeline for systematic review and meta-analysis workflow.

A comprehensive toolkit for enhancing MLLM spatiotemporal reasoning in VideoQA tasks.

Collaboration infrastructure for coordinating autonomous research agents in scientific research.

A multi-domain benchmark for natural-language-to-PDDL specification construction with planner verification.

A suite of automated checkers and audit prompts for formal math datasets and Lean theorem proving.

Framework for multi-agent LLMs to evolve compact symbolic communication protocols.

Dataset and code for evaluating and improving LLM agent abstention decisions.

A dynamic representation editing framework for steering LLM reasoning trajectories toward truth.

A representation-level auditing framework for certifying visual unlearning in Vertical Federated Learning.

Project webpage and code for guidelines on rotation actions in deep RL.

Mamba-based network for spatial reverse attention in pathological liver MRI segmentation.

A threshold-free KV cache compression library for reducing LLM inference memory consumption.

Head-aware KV cache management system for efficient long-context LLM serving.

Benchmark and framework for selective multimodal LLM unlearning and privacy erasure.

A self-evolving agent framework for hardware design as repository-level code evolution.

Agentic framework for reconstructing 3D geometry and 4D multi-object interactions from monocular video.

A scalable training framework for physically consistent embodied video generation.