Article
ALTK-Evolve: How learned memory tops the τ²-bench leaderboard on customer service tasks without a frontier model
ibm-research
• • 20
Ayhan Sebin
ayhansebin
AI & ML interests
None yet
Recent Activity
published an article about 12 hours ago
ALTK-Evolve: How learned memory tops the τ²-bench leaderboard on customer service tasks without a frontier model liked a model 13 days ago
ibm-research/ttm-r3 published an article about 1 month ago
Your Agent Aced the Task. Will It Do It Again?Organizations
published an article about 12 hours ago
liked a
model 13 days ago
published an article about 1 month ago
Article
Your Agent Aced the Task. Will It Do It Again?
ibm-research
• • 122 upvoted an article about 1 month ago
Article
IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly license
ibm-research
• • 60 published an article about 1 month ago
Article
IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly license
ibm-research
• • 60 commented on Real-Time Intelligence with IBM Time Series Models on Confluent about 2 months ago
If you want to try it on your own topics, sign up here: https://events.confluent.io/early-access-flink-features
upvoted an article about 2 months ago
Article
Real-Time Intelligence with IBM Time Series Models on Confluent
ibm-research
• • 53 published an article about 2 months ago
Article
Real-Time Intelligence with IBM Time Series Models on Confluent
ibm-research
• • 53 published an article 2 months ago
Article
How Much Memory Does Your Agent Actually Need?
ibm-research
• • 83 liked a
model 3 months ago
upvoted 2
articles 3 months ago
Article
Security incident disclosure — July 2026
system
• • 807 Article
Model Routing Is Simple. Until It Isn’t.
ibm-research
• • 95 upvoted an article 4 months ago
Article
ScarfBench: Benchmarking AI Agents for Enterprise Java Framework Migration
ibm-research
• • 27 updated a
collection 4 months ago
liked a
Space 4 months ago
commented on Build real agentic apps using CUGA: two dozen working examples on a lightweight harness 4 months ago
Sovereign, Open, Self-Learning Agents: https://github.com/cuga-project/cuga-agent.
We look forward to collaborating and welcoming your contributions to help advance open-agent community.
upvoted an article 4 months ago
Article
Build real agentic apps using CUGA: two dozen working examples on a lightweight harness
ibm-research
• • 40 liked a
Space 4 months ago
liked a
dataset 4 months ago