Easy-to-use LLMs as a single executable file that work across various hardware and OSs.

A tool to bring modern web browsing capabilities to PowerPC Macs

A multi-agent framework for multimodal medical reasoning in Indic languages.

A framework for specifying agent coordination based on message sequence charts to generate deadlock-free agent programs.

Python library for unit circle based computing and variational phasor circuits.

Mixtures of SubExperts for Large Language Continual Learning using compositional routing.

Tool and benchmark for detecting systematic misalignments in MLLM-generated captions.

Agentic scientific computing tool for automating Lattice QCD research workflows.

A minimal interpretable architecture for zero-shot reconstruction of chaotic and cyclic dynamical systems.

An artifact-centered workbench and executable benchmark for LLM agents in structural engineering workflows.

A Python library for implementing Copy-on-Write scoring to evaluate LLM agents within application environments.

A multi-modal orchestration framework for semantic audio-driven humanoid whole body control.

Learned initialization via gated optimization to alleviate convergence failures in Physics Informed Neural Networks.

An open standard for connecting LLMs to external tools and data sources.

A structural security monitor analyzing control-flow and data-flow graph diffs to prevent AI agent sabotage.

An open-source tool layer for reliable, semantic computer-use agents on desktop applications.

Iterative RAG framework for generating scenario scripts for autonomous driving validation.

Multi-Agent System Process Reward Model for improving inference-time search in MAS.

A parameter-efficient convolutional network for handwritten Devanagari recognition.

A framework for learned adaptive memory management for LLM agents using a contextual bandit approach.