IBM Research

company

https://research.ibm.com/

AI & ML interests

Enterprise AI and ML, Foundation Models, Responsible AI

Recent Activity

mustansarfiaz updated a model about 3 hours ago

ibm-research/GeoVLM-R1

samimarreed updated a Space about 12 hours ago

ibm-research/cuga-policy-playground

sanborse authored a paper 5 months ago

GneissWeb: Preparing High Quality Data for LLMs at Scale

View all activity

Papers

DRACO: Fine-Grained Credit Assignment with Dynamic Rubrics for Long-Horizon Agent Training

Evoflux: Inference-Time Evolution of Executable Tool Workflows for Compact Agents

View all Papers

Articles

Ensemble forecasting with IBM Granite time-series models

ALTK-Evolve: How learned memory tops the τ²-bench leaderboard on customer service tasks without a frontier model

Your Agent Aced the Task. Will It Do It Again?

IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly license

Real-Time Intelligence with IBM Time Series Models on Confluent

How Much Memory Does Your Agent Actually Need?

Thinking of ACE? We Can Do It with Fewer Tokens

Model Routing Is Simple. Until It Isn’t.

ScarfBench: Benchmarking AI Agents for Enterprise Java Framework Migration

Build real agentic apps using CUGA: two dozen working examples on a lightweight harness

Beyond LLMs: Why Scalable Enterprise AI Adoption Depends on Agent Logic

ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for Agentic Enterprise IT Tasks — by Artificial Analysis and IBM

The Open Agent Leaderboard

Inside VAKRA: Reasoning, Tool Use, and Failure Modes of Agents

ALTK‑Evolve: On‑the‑Job Learning for AI Agents

IBM and UC Berkeley Diagnose Why Enterprise Agents Fail Using IT-Bench and MAST

AssetOpsBench: Bridging the Gap Between AI Agent Benchmarks and Industrial Reality

CUGA on Hugging Face: Democratizing Configurable AI Agents

View all articles

ibm-research 's Spaces 19

BlueBench Leaderboard

An open-source benchmark for enterprise use cases.

CUGA Policy Playground

Two live CUGA agents, with and without policies

CUGA Apps

Explore and launch CUGA apps from a single dashboard

Risk Atlas Nexus

Evaluate AI risks with common risk taxonomies

VAKRA Leaderboard

Explore and submit AI agent benchmarks on VAKRA

ScarfBench

Java framework migration

BPO Benchmark Evaluation

Run BPO recruiting benchmark with CUGA SDK

Ai Atlas Nexus Capabilities

Intent to AI Capability with AI Atlas Nexus

AssetOpsBench

Run a quick benchmark of web asset performance

ITBench-Lite-Space

Develop and run interactive code notebooks with JupyterLab

CUGA Agent

Configurable Generalist Agent, leader in AppWorld Benchmark

3DGrid-VQGAN Demo

SMILES to 3D grid and reconstruction comparison demo

AssetOpsBench

Evaluating Autonomous AI Agents for Industry 4.0 Tasks

SMI-TED-demo1

Generate embeddings from SMILES strings or CSV files

Fm4m Eval Demo

FM4M-demo2

Generate and analyze molecular structures

FM4M-demo1

Generate and analyze molecular structures

README

Biomed.sm.mv Te 84m

Prediction task tests for biomed-multi-view models