Beyond Memory: A Templated Substrate for Heterogeneous Collaborative Knowledge Work with LLM Agents
Quick Answer
This paper shows that The llm-wiki-memory-template introduces a structured approach for collaborative knowledge work using LLM agents, preserving both successful and failed attempts in research.
Quick Take
This append-only wiki system addresses the issue of lost negative results, enhancing multi-human, , and multi-domain collaboration. Case studies demonstrate its effectiveness in maintaining research integrity and improving outcomes.
Key Points
- Introduces -wiki-memory-template for preserving both successful and failed research outcomes.
- Append-only wiki structure addresses the loss of negative results in collaborative work.
- Supports multi-human, multi-AI-agent, and multi-domain collaboration through distinct architectural elements.
- Case studies show improved research integrity and accurate reporting of experimental results.
- Highlights sociotechnical properties like failure-path preservation and agent honesty.
DeepSignal Analysis
What happened
The llm-wiki-memory-template introduces a structured method for collaborative knowledge work using LLM agents, focusing on preserving both successful and unsuccessful research attempts. This append-only wiki system aims to address the common issue of lost negative results, thereby enhancing collaboration across multiple human and AI agents in various domains.
Key evidence
- The llm-wiki-memory-template is designed to maintain a record of both successful and failed research attempts, which are often excluded from traditional publications.
- Three case studies demonstrate the effectiveness of the llm-wiki-memory-template, including a solo research lineage that preserved abandoned iterations.
- The append-only nature of the wiki addresses the problem of negative-result loss, which is structurally unresolvable in conventional publication and code-sharing practices.
Why it matters
This approach could significantly improve the integrity of research by ensuring that both successes and failures are documented, potentially reducing the repetition of past mistakes. By fostering a more transparent collaborative environment, it may enhance the overall quality of knowledge work across different fields and among various agents.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Research projects, educational efforts, and adjacent knowledge work accumulate findings, decisions, and reasoning that future collaborators rarely recover. The parts most useful to that work, including dead ends and walked-back claims, are routinely excluded from publications and shared code; future researchers re-attempt the same failures because no record survives. LLM coding agents are common participants but hold no persistent memory across sessions, and retrieval-augmented generation over raw sources does not compound. The llm-wiki pattern (Karpathy, 2026; tonbi, 2026) addresses this by inserting an LLM-maintained, interlinked wiki between raw sources and the agent. We present llm-wiki-memory-template, a reusable, agent-aware instantiation, and argue it is a substrate for heterogeneous collaborative knowledge work along three axes (multi-human, multi-AI-agent, multi-domain) with each axis supported by a distinct architectural element of the template (§4). The wiki is append-only by convention, which preserves what did not work alongside what did, addressing a negative-result loss problem that publications and code-sharing structurally cannot solve. Three deployed case studies and one design report cover the axes individually: a solo research lineage that preserves abandoned iterations; a two-author project whose retroactive audit revised two prior experiments' claimed 20-of-20 coverage down to 14 and 12 evidence-based answers, then to 18 and 18 after a fix, with the failure path preserved across the artifact; an in-progress multi-agent deployment reported as a design; and a cross-domain educational variant. We name failure-path preservation, agent honesty, and appropriation as cross-cutting sociotechnical properties of the artifact, not only of its technical mechanisms.
| Comments: | 15 pages, 3 figures, 1 table |
| Subjects: | Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Digital Libraries (cs.DL) |
| ACM classes: | H.5.3; I.2; D.2 |
| Cite as: | arXiv:2607.24759 [cs.AI] |
| (or arXiv:2607.24759v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.24759 arXiv-issued DOI via DataCite |
Submission history
From: Christopher Sweet [view email]
[v1]
Fri, 29 May 2026 15:34:14 UTC (22 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.