Buffer of Thoughts
Thought-augmented reasoning with LLMs. NeurIPS 2024.
At a glance
- Type
- Thought-augmented reasoning
- Tier
- T3
- Created
- 2024-06-06 (arxiv 2406.04271 submitted June 6 2024; NeurIPS 2024 Spotlight)
- Latest release
- not applicable — not OSS
- License
- not applicable — not OSS
- GitHub
- not applicable — no GitHub repo
- Pricing
- not applicable — not commercial
- Funding
- not applicable — not commercial
Taxonomy
- storage
- kv
- retrieval
- similarity
- persistence
- session
- update
- append-only
- unit
- chunk
- governance
- inspectable
- conflict
- llm-arbitrate
When to use
Optimised for: not applicable - research paper
Anti-fit: not applicable - research paper
Pros & cons
Pros
Thought-augmented reasoning — caches and reuses high-level thoughts across queries; NeurIPS 2024.
Cons
Buffer maintenance overhead; thought-quality bounds reuse value.
Claims & capabilities
Stores high-level thought-templates distilled from problem-solving; buffer-manager dynamically updates the meta-buffer at 12% of the compute of competing methods. Headline: +11% on Game of 24, +20% on Geometric Shapes, +51% on Checkmate-in-One over previous SOTA; baseline: previous SOTA prompting/reasoning methods; primary dataset: 10 reasoning-intensive tasks including Game of 24, Geometric Shapes, Checkmate-in-One.
Technical surface
- API surface
- not applicable — research paper
- Backend storage
- not applicable — research paper
- Deployment
- not applicable — not a deployable product
- Embedding model
- not applicable — research paper
- Multi-tenancy
- not applicable — research paper
- MCP
- not applicable — research paper, no deployed product
- A2A
- not applicable — research paper, no deployed product
- OpenTelemetry
- not applicable — research paper, no deployed product
Similar systems
Other recent method papers — theorized, no distinct product in the catalog, ranked by inbound references.
- Compressive Transformer T3
Maintains recent states in full resolution while compressing older memories with learned compression functions. DeepMind.
- MemGPT v2 / agent-tools T3
Already in catalog as the foundational MemGPT paper. Note: Letta is the productionised successor (cross-listed).
- Transformer-XL T3
Extends context through segment-level recurrence + caching of hidden states from prior segments. Foundational long-context architecture.
- Generative Agents T3
Park et al. — landmark agent-simulation paper. Reflection + memory stream + retrieval enable believable agent behavior.
- MemoryBank T3
Enhances LLMs with long-term memory. Early influential paper.
- Reflexion T3
Language agents with verbal reinforcement learning.