Player-centric basketball event understanding dataset and reasoning framework.

A benchmark for evaluating the safety and reliability of autonomous LLM agents.

Research paper investigating attention mass, function token anchoring, and retrieval limits in LLMs.

An optimized hypergraph-based RAG approach for better fact extraction and chunk retrieval.

A framework for sign-guided merging and adaptation in multimodal continual instruction tuning.

A benchmark for evaluating LLM coding capabilities with verifiable rubrics.

A tool for creating and experimenting with CSS gradients.

Synthetic benchmarking tool for evaluating computer and phone performance.

A low-latency streaming framework for interaction-aware temporal memory adaptation in HOI video generation.

A hybrid architecture combining Qwen3-4B and Foundation-Sec-8B for effective malware detonation report interpretation.

A framework for producing high-quality full-length music from lyrics, text, and musical attributes.

System for multi-perspective news generation and bias-aware evaluation using LLMs.

Embeddable email and document builders for apps

An AI-driven legal question answering system tailored to the Indian legal system using RAG.

A theory and agent framework focusing on value-aware forgetting as a learning mechanism.

A survey and framework for verifying and auditing agentic AI in critical engineering domains.

Python interface for the SCIP optimization suite

Verilog simulator and synthesizable tool for hardware design verification.

Open-source framework for RTL synthesis.

Research paper analyzing shared capacity and redundancy in multi-task learning