AgentFold
https://arxiv.org/abs/2510.24699
Long-horizon web agents with proactive context management.
At a glance
- Type
- Proactive context management
- Tier
- T4
- Created
- 2025-10
- 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
- agentic
- persistence
- session
- update
- consolidation
- unit
- summary
- governance
- inspectable
- conflict
- llm-arbitrate
When to use
Optimised for: not applicable - research paper
Anti-fit: not applicable - research paper
Pros & cons
Pros
Folded-context agent — compresses old context into summary memory.
Cons
Preprint; folding-quality bounds long-horizon recall.
Claims & capabilities
Proactive context management via folding inspired by retrospective consolidation; multi-scale granular condensation to deep consolidation. Headline: AgentFold-30B-A3B reaches 36.2% on BrowseComp and 47.3% on BrowseComp-ZH, surpassing DeepSeek-V3.1-671B (30.0% on BrowseComp) and OpenAI o4-mini; baselines: DeepSeek-V3.1-671B, GLM-4.5-355B, OpenAI o4-mini; primary datasets: BrowseComp, BrowseComp-ZH, WideSearch, GAIA.
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.