Capability & product
Frontier capability releases, new product surfaces, model launches.
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
arXiv cs.CL (Computation & Language): Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregressive LLMs by enabling non-autoregressive text generation. However, their practical deployment rema...
7+ days agoAgensh: Scaling Organizational Intelligence to 1,024 Agents
arXiv cs.CL (Computation & Language): A multi-agent system can reduce latency on complex tasks by executing work concurrently. Several pioneering harness frameworks support multi-agent systems. However, the scalability of current multi-ag...
7+ days agoSpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue
arXiv cs.CL (Computation & Language): Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish who said what, whom each statement concerns, ho...
7+ days agoBeyond Repeated Sampling: Learning Search Policies for LLM Reasoning
arXiv cs.CL (Computation & Language): Large language models increasingly tackle hard reasoning problems by spending more test-time compute, yet the dominant strategy remains naive repeated sampling: draw many independent solutions and hop...