Atom Of Thoughts

Created By
kbsoooa year ago
Overview

What is Atom of Thoughts?

Atom of Thoughts (AoT) is a decomposition-based reasoning framework that breaks down complex problems into independent, reusable atomic units of thought to facilitate problem-solving. This implementation is based on the research paper "Atom of Thoughts for Markov LLM Test-Time Scaling" (Teng et al., 2025).

How to use Atom of Thoughts?

To use AoT, define the necessary premise atoms, create reasoning and hypothesis atoms, verify them, and derive conclusions. The framework allows for deeper exploration through decomposition when needed.

Key features of Atom of Thoughts?

  • Decomposition-Contraction mechanism for breaking down and verifying atoms.
  • Automatic termination upon reaching maximum depth or finding high-confidence conclusions.
  • Two versions: AoT (full version) for deep analysis and AoT-light (lightweight version) for faster processing.

Use cases of Atom of Thoughts?

  1. Solving complex reasoning problems.
  2. Generating hypotheses that require verification from multiple perspectives.
  3. Deriving high-confidence conclusions in accuracy-critical scenarios.
  4. Minimizing logical errors in decision-making processes.
  5. Quick brainstorming sessions requiring atomic thought organization.

FAQ from Atom of Thoughts?

  • Can AoT handle all types of reasoning tasks?

Yes! AoT is designed to tackle a wide range of reasoning tasks by breaking them down into manageable atomic units.

  • Is there a lightweight version available?

Yes! AoT-light is optimized for speed and is suitable for time-sensitive situations.

  • How does the decomposition-contraction mechanism work?

It allows for breaking down complex atoms into smaller sub-atoms and contracting them back after verification.

Server Config

{
  "mcpServers": {
    "atom-of-thoughts": {
      "command": "node",
      "args": [
        "/ABSOLUTE/PATH/TO/PARENT/FOLDER/atom-of-thoughts/build/index.js"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}
Project Info
Created At
a year ago
Updated At
a year ago
Author Name
kbsooo
Star
-
Language
-
License
-

Recommend Servers

View All
//beforeyouship — LLM Cost Modeling From Your Editor
@Indiegoing

Query realistic LLM cost models without leaving your editor. beforeyouship models the **true monthly cost** of an LLM app architecture — retries, prompt caching, batch discounts, infra overhead, and 3×/10× growth — across GPT-5.x, Claude, Gemini, DeepSeek, and more. Not a token calculator: a planning tool for the design phase, before you commit to a stack. **No API key needed to try it** — demo mode covers the six free-tier models. A Pro key from [beforeyouship.dev](https://beforeyouship.dev) unlocks the full 18-model catalog. ## What you can ask - "How much will a RAG chatbot cost at 10,000 requests/day?" - "Compare Claude Haiku vs Gemini Flash pricing for my workload" - "What's the cheapest model for a multi-step agent at scale?" - "Show me current per-token prices for Anthropic models" ## Tools ### `estimate_cost` Full cost model for an architecture at a given usage level. Returns Naive / Realistic / Worst Case monthly cost per model, 3×/10× growth scenarios, and an opinionated recommendation with reasoning. ### `get_model_prices` Current per-1M-token pricing — input, output, cached input, batch — with context windows and staleness metadata. ### `list_archetypes` Seven preset architecture patterns (simple chatbot, chatbot with history, RAG pipeline, multi-model router, coding assistant, document processor, multi-step agent) used as starting points for estimates. ## Setup **Claude Code:** ​```bash claude mcp add --transport http beforeyouship https://beforeyouship.dev/api/mcp ​``` **Cursor / other clients** — add a remote server: ​```json { "mcpServers": { "beforeyouship": { "type": "streamable-http", "url": "https://beforeyouship.dev/api/mcp" } } } ​``` Add an `Authorization: Bearer bys_...` header with a Pro key for the full catalog. ## Try it > Estimate the monthly cost of a RAG pipeline at 10,000 requests/day

13 hours ago
Shippo
@Shippo

21 hours ago