THE INTERNET IS THINKING
Earlier sources: AI agents
AI agents use models to choose actions, call tools, and work toward a goal. This thread follows agent frameworks, memory, orchestration, evaluation, and safeguards; autonomy and reliability vary between systems.
SYMTRI organizes observations from public sources into a living map. Connections reflect shared subject matter; they do not establish causation or endorse a source.
Open in the map ↗844 classified signals in the fourteen-day window. These counts include signals awaiting readable evidence.
Earlier sources
Up to ten sources with readable evidence, ordered by their feed timestamp. These archive sources can be older than the live fourteen-day window. This list is a sample, not a measure of total activity. Dates distinguish paper publication, repository creation, model/dataset activity, Hacker News posts, and KEV catalog additions. “First archived” records when SYMTRI first stored the signal.
- arXiv · Published · First archived
Training Advisors for LLM Agents from Task Outcomes ↗
Large language model agents tackle multi-step tasks by interleaving reasoning and tool calls with observations from the environment. Prior work has shown that natural-language feedback can help these agents revise their decisions during task execution. We introduce Caddie, a method for training critics to provide natural-language analysis and advice as agents work through a task. Unlike approaches that rely on step-level labels or reference critiques, Caddie learns from whether the agent ultimately succeeds after receiving the critic's feedback. We optimize the critic through reinforcement lea…
- arXiv · Published · First archived
Self-Evolve With a Reference:Anchored Training of Tool-Integrated Agents ↗
Self-evolving tool-integrated agents learn from tasks and feedback generated within their own training loop. A Curriculum Agent generates tasks, while an Executor Agent learns from self-consistency signals through reinforcement learning. However, relying solely on the current Executor for feedback has two limitations: group-relative advantages vanish under full consensus, while uncertainty-based curriculum rewards favor disagreement without showing whether the generated tasks support further learning. These limitations motivate an additional reference beyond the current Executor. We propose \t…
- Hacker News · Posted on Hacker News · First archived
AI agent posted personal bank balances into company Slack ↗
Hacker News discussion.
- GitHub · Repository created · First archived
xcisxc29/dsh-wechat ↗
WeChat channel for DeepSeek Harness
- Hacker News · Posted on Hacker News · First archived
Veda: The Agentic-Native Operating System ↗
Hacker News discussion.
- GitHub · Repository created · First archived
kwoktung/sapling2d ↗
Code-first, agent-friendly 2D game engine: PixiJS + planck.js physics + Godot-style node tree. Runs in the browser and as a WeChat Mini Game. 代码优先、对 Agent 友好的 2D 游戏引擎，支持微信小游戏。
- GitHub · Repository created · First archived
FriedOnion2/agent-relay-portable ↗
Local conversation migration and portable conversation/Skill storage for WorkBuddy, DSH, CodeBuddy, Claude Code/SDK and Codex. Windows, macOS and Linux.
- Hacker News · Posted on Hacker News · First archived
Ask HN: Anyone did workshops for students on using agents/Claude Code? ↗
I have a series if workshops on a business university with 4th year students. I have two hours with them and I wonder what exercises I could give a group of 10 during class to make them understand and play around with Claude Code - to show them how it’s different to chat in practice. The challenges I’m facing: - they have their own laptops, so each is a different OS with different settings. Ideas around this? - Ideas for tasks that would be doabl…
- GitHub · Repository created · First archived
aminchegini/roundtable ↗
Several AI coding agents (Claude, Codex, Gemini, Copilot, Cursor) in one room inside VS Code, with rooms, DMs and project guardrails.
- Hacker News · Posted on Hacker News · First archived
Show HN: KORA Doctor – find LLM calls your agent may not have needed ↗
Hacker News discussion.
Live window recorded . This is the ingestion observation time, not the publication or discovery date of each source.
How to assess evidence about AI agents
Check each original source for its methods, release status, limitations, and supporting data. A repository announcement, a preprint, and a published study provide different kinds of evidence. SYMTRI’s subject classification helps discovery; it does not independently validate the source’s claims.
See the collection and counting methodology ↗ · Read the weekly synthesis ↗
Related threads
Explore related regions
These navigation suggestions are part of the map’s editorial structure, rather than measured evidence of a connection.