Training LLM agents for extremely long-horizon tasks remains an open challenge. Most agent training pipelines struggle with extended-duration trajectories. Context gets lost, rewards are sparse, and the learning signal degrades over long sequences. KLong tackles this with a
AGENTS
-
Encouragement to build with smolagent’s CodeAgent and propose improvements
By
–
Nice work!
Don't hesitate to build a bit with smolagent's CodeAgent if you're interested in code-actions agents, I'm less involved now but would still love to look at what improvements you could propose! -
Industrial AI Requires Governed Agents Not Autonomous Systems
By
–
In manufacturing, a 90% success rate is a failure. AI agents struggle with broad questions that require 100% deterministic answers, often leading to hallucinations.
— Lucian Fogoros (@fogoros) 20 février 2026
The future of Industrial AI isn't autonomous chaos; it's governed agents. @HighByteInc Intelligence Hub version… pic.twitter.com/j9A4gHp4lkIn manufacturing, a 90% success rate is a failure. AI agents struggle with broad questions that require 100% deterministic answers, often leading to hallucinations. The future of Industrial AI isn't autonomous chaos; it's governed agents. @HighByteInc Intelligence Hub version
-
Structuring Skills for AI Agents with Skill Graphs
By
–
Skill Graphs > SKILL .md Everyone's talking about skills for AI agents. But almost nobody is talking about how to structure them. Right now, the default approach is simple. You write one skill file that captures one capability. A skill for summarizing. A skill for code review.
-
Dynamic Multi-Agent Systems with Population Growth and Adaptation
By
–
Interesting new work on dynamic multi-agent systems. Traditional multi-agent RL assumes a fixed number of agents. But real-world scenarios often involve populations that grow, split, or shrink dynamically, such as cell division, team expansion, or organizational scaling.
-
Five Critical AI Agent Mistakes Costing Businesses Millions
By
–
The 5 AI Agent Mistakes That Could Cost Businesses Millions #AI #agents are set to #transform how businesses operate, but many organizations are walking into this shift unprepared. This article explores the five biggest #mistakes #leaders will make with #AIagents, from #data
-
OAuth Token Optimization and Cache Miss Challenges in Agent Harnesses
By
–
I think this is why using the OAuth token with other agent harnesses isn't great – they do not do the same optimizations that Anthropic is doing (like figuring out all the compaction edge cases). My hunch is that those unoptimized harnesses cause cache misses, which end up
-
Multi-Agent AI DAO Prior Art: Autonomy Rights Since 2017
By
–
The "Multi-Agent AI DAO" Prior Art—pioneered by http://
MONTREAL.AI in 2017—forever enshrined every AI agent’s right to autonomy. Watch the Proof: https://
youtu.be/Y4_6aZbVlo4?si
=UeHJeINn6NMxG_Cv
… Historical Comparables to the 2017 hashtag#PriorArt “Multi-Agent AI DAO” 1. -
Agentic RAG Chatbot for Secure Medical Data Retrieval
By
–
Building an agentic RAG chatbot for closed source medical data retrieval with high sensitivity for data privacy. What tools do you suggest?
-
Server-Side Tool Call Schema Validation Best Practices
By
–
Not obvious, no. In fact OpenAI already does it correctly server side. There's no reason for a provider to not do this. Or to at least document that tool call schemas are merely a suggestion.