IBM Research
PulseAugur coverage of IBM Research — every cluster mentioning IBM Research across labs, papers, and developer communities, ranked by signal.
- subsidiary of IBM 100%
- developed Granite Embedding Multilingual R2 95%
- developed Vakra 95%
- developed by ScarfBench 95%
- developed by Vakra 90%
- developed ScarfBench 90%
- used by Vakra 60%
- affiliated with Vakra 60%
- competes with Inkling 50%
- affiliated with DeepInfra 50%
- competes with DeepInfra 50%
- affiliated with Inkling 50%
7 day(s) with sentiment data
-
IBM, Databricks, and ElevenLabs announce AI advancements · 3 sources tracked
IBM Research is developing Praxis, a system designed to integrate policy enforcement directly into the AI stack. Meanwhile, Databricks and Mosaic AI have introduced UCDataset to enhance GPU training performance on Unity…
-
AI Agent Communication Protocol A2A Donated to Linux Foundation
The Agent2Agent (A2A) Protocol, an open standard for AI agent communication, has been donated to the Linux Foundation and is maintained by a technical steering committee including representatives from major tech compani…
-
Peking University and IBM Research Define Modern Analog Computing
Researchers from Peking University, led by Sun Zhong, have collaborated with IBM Research Europe to define modern analog computing. Their review centers on three core primitives that form the basis of this computing par…
-
NVIDIA optimizes AI performance, IBM Research looks back at 70 years of algorithmic innovation
NVIDIA has published details on optimizing Tensor Core performance for machine learning applications, specifically focusing on interatomic potentials for silicon structural energies. Separately, IBM Research is highligh…
-
IBM Research launches Java migration leaderboard and advances quantum computing
IBM Research, in collaboration with Kaggle, has introduced a leaderboard focused on migrating Java frameworks. This initiative aims to track and encourage progress in the complex task of transitioning Java-based systems…
-
IBM Research: PyTorch now supports new AI accelerators
IBM Research has announced that PyTorch is now compatible with a new class of AI accelerators. This development aims to broaden the accessibility of AI development by allowing a wider range of hardware to be utilized fo…
-
Cornell and IBM Research share KV cache across recursions
Researchers from Cornell University and IBM Research have developed a method to share the KV cache across recursive calls in language models. This technique allows the initial recursive step to build a memory, which sub…
-
Tools convert PDFs to Markdown for LLMs, aiding RAG and data extraction
Two articles discuss methods for converting PDF documents into Markdown format, a structure that LLMs can better process for tasks like retrieval-augmented generation (RAG) or prompt inclusion. The first article compare…
-
Self-hosted RAG solutions gain traction for enterprise data security
Enterprises are increasingly adopting Retrieval-Augmented Generation (RAG) to access internal sensitive data, but security concerns necessitate self-hosted solutions. RAG systems can be vulnerable to attacks that reveal…
-
New method tackles AI agent inconsistency, improving reliability
Researchers have developed a new method to address the inconsistency problem in AI agents, where a task that succeeds once may fail on subsequent attempts. This issue, often masked by average performance metrics, is par…
-
JetBrains, Thinking Machines, and IBM Research launch new AI models and insights · 3 sources tracked
JetBrains has launched Mellum2, a 12 billion parameter expert mixture model, detailed on Hugging Face. Separately, Thinking Machines has introduced Inkling, also featured on Hugging Face. IBM Research has published insi…
-
IBM Research unveils Vakra analysis tool and Granite multilingual embedding model
IBM Research has released two new AI models: Vakra, which analyzes agent reasoning, tool usage, and failure modes, and Granite Embedding Multilingual R2, an open-source multilingual embedding model with a 32K context wi…
-
AI research on agent memory and LLM benchmarks highlighted
Two Mastodon posts highlight recent research from Hugging Face blogs concerning AI agents and LLM benchmarks. The first post discusses the memory requirements for AI agents, referencing research from IBM Research on ALT…
-
IBM Research open-sources AI policy schema for GenAI
IBM Research has released an open-source policy schema designed for generative AI applications. This schema, presented in YAML format, along with a synthetic data pipeline, aims to help enforce content rules throughout …
-
Hugging Face Spotlights AI Advancements in DPO, New Models, and Routing
Hugging Face is highlighting several recent advancements in AI, including a blog post on Direct Preference Optimization (DPO) which enables language models to act as reward models. Another post discusses newer models th…
-
IBM Research explores agentic development bottlenecks
IBM Research developer advocate Markus Eisele is exploring agentic development, focusing on the crucial aspects of creating effective system specifications and ensuring quality assurance for AI-generated outputs. His wo…
-
IBM Research unveils Vakra and Granite Embedding Multilingual R2 models
IBM Research has released two new models via Hugging Face: Vakra, a benchmark for analyzing agent reasoning, tool usage, and failure modes, and Granite Embedding Multilingual R2, an open-source multilingual embedding mo…
-
Agent memory boosts some AI models, but not others, IBM Research finds
IBM Research conducted tests on agent memory across eight AI models, observing varied performance improvements. The largest model tested showed no gains, while an 117B parameter model saw a 16-point increase in performa…
-
Hugging Face and IBM Research explore AI agent memory needs
A blog post from Hugging Face, co-authored by IBM Research, explores the memory requirements for AI agents. The article, titled "How Much Memory Does Your Agent Actually Need?", delves into the specifics of memory capac…
-
Hugging Face blog posts cover AI openness, model updates, and routing · 6 sources tracked
Hugging Face is publishing a series of blog posts covering various AI topics. These include discussions on the importance of openness in AI and cybersecurity, the addition of MCP tools to Reachy Mini, and an analysis of…