Introducing DSGym: A New Framework for Evaluating Data Science Agents
DSGym is a new framework for evaluating and training data science agents, offering standardized solutions.
#Language Models (14)
Accelerate Attention with FlashAttention-3: New Capabilities and Performance
FlashAttention-3 significantly accelerates attention in AI models, achieving 1.2 PFLOPS with FP8 and improving GPU performance.
Run Isaac 0.1 on Replicate and Discover More
Isaac 0.1 is a new model for visual perception and OCR by Perceptron AI.
Explore the capabilities of Nano Banana Pro for image creation
Nano Banana Pro impresses with its capabilities in image and text creation.
IBM Releases Granite 4.0 for Fast Language Model Usage
IBM has launched Granite 4.0 — new open-source language models for business.
Announcing Replicate's remote MCP server for applications
Replicate announced a remote MCP server for applications, simplifying access to APIs.
Microsoft Unveils Harrier-OSS-v1: New Multilingual Models Released
Microsoft has released Harrier-OSS-v1, a new family of multilingual models achieving SOTA results.
·
3 просмотров
Salesforce AI Unveils VoiceAgentRAG: Boosting Voice RAG Retrieval Speed
Salesforce AI introduces VoiceAgentRAG, an architecture that significantly boosts voice query processing speed.
·
11 просмотров
Hugging Face Launches TRL v1.0: A Unified Post-Training Stack
Hugging Face has launched TRL v1.0, setting a new standard in LLM post-training.
·
28 просмотров
Google AI Launches Veo 3.1 Lite for Low-Cost High-Speed Video
Google has launched Veo 3.1 Lite, a model for low-cost video generation.
·
27 просмотров
Liquid AI Releases LFM2.5-350M: A Compact 350M Parameter Model
Liquid AI has released LFM2.5-350M, a model with 350M parameters trained on 28T tokens, showcasing high efficiency and intelligence density.
Evaluating Invoice Data with LLM as a Judge
Learn how to evaluate invoice data using LLM as a judge to enhance AI accuracy.
·
15 просмотров
How a PhD Student Created the Attention Mechanism in Neural Networks
Discover how a PhD student created the attention mechanism in neural networks by solving practical translation problems.
·
13 просмотров
Introducing Gemma Scope 2 for Analyzing Language Model Behavior
Gemma Scope 2 has been announced - a new toolkit for analyzing language models that will help researchers understand their behavior.
·
51 просмотров