Track synthesis, materials and catalysis as it is published.
Search papers, labs, and topics across Lattice.
The most-read papers, busiest labs, and emerging topics - updated every morning.
We read everything so you don't have to. One Lattice Chemistry email, no fluff.
Omni-modal foundation models achieve self-improvement via RL compute scaled to 1M context, leveraging asynchronous training and agentic grading for enhanced intelligence.
β This work provides a scalable framework for achieving model self-improvement through reinforcement learning, paving the way for more capable and adaptable AI systems.
Irrelevant text distractors cripple mixed-modality retrieval performance more than images, revealing a text modality preference that degrades performance in a V-shaped curve.
β This work provides a critical understanding of modality biases in multimodal retrieval, paving the way for more robust and equitable information access across diverse data types.
ACT3 reveals that integrating carbon cost into silicon design can transform how we approach semiconductor architecture, making sustainability a key performance metric.
β This tool could fundamentally shift the semiconductor industry towards prioritizing sustainability, influencing both research directions and practical design choices.
PatchWrite achieves a flawless preservation of manuscript integrity, maintaining 100% accuracy in edits while traditional methods fail completely.
β This advancement could fundamentally change how automated manuscript editing systems ensure the reliability and accuracy of AI-generated content.
User-defined keyword spotting can now achieve state-of-the-art accuracy and adapt to individual speakers with minimal data and parameter updates, thanks to a novel dual-stage matching and multi-modal enrollment framework.
β Enables more personalized and efficient voice interfaces on edge devices, reducing reliance on large datasets and compute-intensive models.
Rolling 7-day digest, refreshed daily after ingestion at 06:00 UTC