IBM Research
PulseAugur coverage of IBM Research — every cluster mentioning IBM Research across labs, papers, and developer communities, ranked by signal.
7 day(s) with sentiment data
-
Hugging Face blog posts cover AI agent internals, inference providers, and async processing · 3 sources tracked
This cluster highlights three technical blog posts from Hugging Face, each focusing on a different aspect of AI infrastructure and research. The first post delves into the internal workings of Vakra, an AI agent, examin…
-
Stale documents in RAG systems pose significant risks, study finds
A recent study conducted by Emory University and IBM Research investigated the impact of stale documents on retrieval-augmented generation (RAG) systems. The experiment revealed that outdated information in a RAG system…
-
IBM unveils world's first sub-1nm chip technology with stacked transistors
IBM has announced a significant advancement in chip technology with its new "sub-1 nanometer" (0.7nm or 7 angstrom) chip process. This breakthrough utilizes a novel "nanostack" architecture, which vertically stacks tran…
-
Hugging Face highlights AI advancements: Transformers v5, DeepSeek V4, and more · 7 sources tracked
Hugging Face is highlighting several recent advancements across the AI ecosystem. These include the release of Transformers v5 for defining AI models, the OpenEnv framework for evaluating tool-using agents in real-world…
-
AI agent waste detector pivots from failure prediction to token-saving
The developer of an AI agent waste detection tool, Clew, initially hypothesized that structural cycles and embedding decay could predict multi-agent failures. However, testing on the MAST-Data dataset yielded poor resul…
-
Hugging Face Blog Posts Cover AI Agents, Inference, and Processing
This cluster highlights three blog posts from Hugging Face, each focusing on a different aspect of AI infrastructure and research. The first post delves into the internal workings of Vakra, an AI agent developed by IBM …
-
AI RAG Architecture Solves Financial Data Ingestion Challenges
This article details a production-ready architecture for Retrieval-Augmented Generation (RAG) systems, particularly for the financial industry where data is complex and unstructured. It emphasizes the critical need for …
-
Hugging Face announces updates including Transformers.js, new leaderboards, and AI tools
Hugging Face has released several updates and new projects across its platform. These include Transformers.js v4 available on NPM, custom CUDA kernels for Codex and Claude, and the Open ASR Leaderboard with new multilin…
-
LlamaIndex and IBM parsers tested for RAG document prep
This article evaluates two open-source document parsers, LitParse from LlamaIndex and Docling from IBM Research, for their effectiveness in preparing documents for Retrieval-Augmented Generation (RAG) pipelines. The eva…
-
Researcher's multi-agent AI failure prediction model fails
A researcher attempted to develop a predictive model for multi-agent AI system failures, hypothesizing that signals like "Loop Pressure" and "Information Gain Decay" could indicate an impending breakdown. The experiment…
-
Frontier AI models fail new IT benchmark, scoring below 50%
A new benchmark, ITBench-AA, has been released to evaluate the capabilities of frontier AI models on enterprise IT tasks, specifically focusing on Site Reliability Engineering (SRE). In initial tests, even the most adva…
-
AI agents gain cross-platform communication with new A2A Protocol v1.0
Microsoft has released version 1.0 of its A2A Protocol, a stable and production-ready standard for cross-platform communication between AI agents. This protocol aims to eliminate the need for custom integration code by …
-
Hugging Face details Holotron-12B, VAKRA, and open-source AI trends
Hugging Face has released details on Holotron-12B, a model designed for high-throughput computer use agents. Additionally, a blog post explores the current state of open-source contributions to Hugging Face as of Spring…
-
AI model evaluations are becoming a costly bottleneck, surpassing training expenses
AI model evaluations are becoming prohibitively expensive, with recent benchmarks costing tens of thousands of dollars and consuming thousands of GPU hours. This high cost is particularly pronounced for agent-based eval…
-
IBM Research integrates vLLM into its RITS Platform for AI development
IBM Research has integrated vLLM, an open-source library for fast LLM inference, into its RITS Platform. This integration aims to enhance the platform's capabilities by leveraging vLLM's efficient processing for large l…