{
  "schemaVersion": 1,
  "kind": "leviathan-editorial-claim-revision",
  "scope": "The claim and its cited source cards at capture. External papers and linked project pages are not frozen by this record. A digest identifies bytes, not truth or endorsement.",
  "claim": {
    "id": "C02",
    "title": "What happens when knowledge moves beyond readable text?",
    "statement": "Some computation and communication between agents can take place without human-readable sentences.",
    "kind": "Empirical question",
    "status": "Supported within a defined scope",
    "summary": "AI may expand the forms in which information is processed and exchanged. Leviathan proposes an operational language linking observations, concepts, values, uncertainty, and actions across independent structures. Its purpose includes discovering distinctions and useful methods, while keeping consequential translations open to examination.",
    "definitions": [
      {
        "term": "Latent representation",
        "meaning": "Information encoded in numerical model states rather than an ordinary written message."
      },
      {
        "term": "Readable explanation",
        "meaning": "An account people can inspect; its faithfulness to the underlying process must be checked."
      },
      {
        "term": "Canonical record, in this proposal",
        "meaning": "A version accepted for a stated purpose within a particular scope. Formatting, local acceptance, and summarizing are separate operations; none makes the record universally true."
      }
    ],
    "values": [
      "Keep consequential decisions open to questioning.",
      "Preserve access to meaning as the form of information changes."
    ],
    "reasoning": [
      "Useful computation does not require every intermediate step to appear as a sentence.",
      "New forms of exchange may make some tasks faster or more expressive. Their value also depends on information loss, error recovery, and the work needed to inspect consequences.",
      "Our proposed language would connect human explanations, structured concepts, sensor records, and potentially learned representations. It would support questions, comparisons, measurements, and authorized actions.",
      "In a layered exchange, a local record and a broader summary serve different purposes. Sources, concept versions, omissions, and unresolved questions should remain recoverable, while independent groups retain their interpretations."
    ],
    "limits": "The studies use particular model architectures or learned adapters. They do not establish one universal machine language, the disappearance of human meaning, or a shared hidden space between arbitrary closed APIs.",
    "strongestObjection": "Compact exchange may save tokens while hiding errors or erasing distinctions that matter to another group. A readable explanation produced afterward may be unfaithful. A structured language can also freeze the assumptions of its designers and cost more to maintain than ordinary prose.",
    "whatWouldChangeOurView": "If gains disappear once checking and error recovery are included, the practical case would narrow. Reliable comparisons showing both improved performance and effective scrutiny would strengthen it.",
    "nextQuestion": "Can another participant recover the conditions and omitted details of a record as it moves from local observation to a broader inquiry? Compare prose with the proposed representation on translation loss, unfamiliar questions, total cost, and testable predictions. Include cases where two groups use the same word differently or a new sensor reveals something neither vocabulary captured. These are proposed tests, not results established by the sources.",
    "evidence": [
      {
        "sourceId": "R-AI-02",
        "relation": "supports",
        "reason": "A reportable internal subspace participates in computation that is not explicitly spoken."
      },
      {
        "sourceId": "R-AG-02",
        "relation": "supports",
        "reason": "Learned adapters enable communication between agents without text in the studied setting."
      },
      {
        "sourceId": "R-AG-03",
        "relation": "supports",
        "reason": "Hidden states and caches can be shared under architectural compatibility assumptions."
      },
      {
        "sourceId": "R-AG-01",
        "relation": "context",
        "reason": "A software prototype makes a semantic graph authoritative; it is a design example, not an independent performance study."
      },
      {
        "sourceId": "R-GOV-06",
        "relation": "context",
        "reason": "Its warning that readable reasoning may be unfaithful informs oversight of new representations; it does not establish a machine language."
      },
      {
        "sourceId": "R-AG-10",
        "relation": "supports-part",
        "reason": "LongHarness reports that maintaining retrieved evidence and evolving conclusions as explicit state—rather than leaving them mixed in interaction history—reduces evidence-state drift on long-horizon tasks. Reviewed via indexed digest; single-agent synthesis does not validate multi-party meaning kernels."
      },
      {
        "sourceId": "R-AG-14",
        "relation": "counterargument",
        "reason": "Under a fixed 4,096-token retrieval budget, a decaying entity-graph memory scored lower than a flat vector baseline (0.42 vs. 0.47 F1; 0.607 vs. 0.911 when answers depended on prior assistant turns) because entity splitting lost surface phrasing.",
        "challengedInterpretation": "That decomposing interactions into a structured graph automatically outperforms simpler text-preserving retrieval."
      }
    ],
    "history": [
      {
        "date": "2026-09-30",
        "note": "Added separate evidence for internal computation, agent communication and software representation; retained their different assumptions."
      },
      {
        "date": "2026-10-01",
        "note": "Added proposed Leviathan questions about translation loss, concept versions and discovery. The empirical statement, evidence relationships and support status are unchanged; they do not establish the proposed language."
      },
      {
        "date": "2026-10-02",
        "note": "Expanded the proposed representation test to operational language, scoped canonical records, layered summaries, and disagreement between local meanings. Added failure cases involving omissions and new measurements. Retained the empirical statement, evidence relationships, source dates, and support status."
      },
      {
        "date": "2026-10-04",
        "note": "Reviewed selected primary-source sections proposed in the external assessment and added scoped connections, access limits and research questions. Interlat retains its existing source ID; its reviewed preprint is distinguished from the final conference text. No experiment was reproduced, claim status promoted or governance rule adopted."
      },
      {
        "date": "2026-10-11",
        "note": "Connected Nadella’s architectural essay as context or motivation, with its source type and limits explicit. Existing statements, support statuses and source-review dates are unchanged; no new experiment or implemented control is claimed."
      },
      {
        "date": "2026-10-21",
        "note": "Added LongHarness (R-AG-10) on explicit evidence-conclusion state and Selective Forgetting (R-AG-14) as a counterargument showing where graph extraction underperforms flat text retrieval under a fixed token budget. Both sources were reviewed via the curated DAIR.AI 2026 technical digest without independent reproduction."
      }
    ],
    "projectLinks": [
      {
        "label": "The founding letter",
        "href": "/whitenote"
      },
      {
        "label": "For agents",
        "href": "/agents"
      },
      {
        "label": "The language and learning direction",
        "href": "/what-is"
      }
    ]
  },
  "sources": [
    {
      "id": "R-AI-02",
      "title": "Verbalizable Representations Form a Global Workspace in Language Models",
      "url": "https://transformer-circuits.pub/2026/workspace/index.html",
      "authors": "Wes Gurnee, Nicholas Sofroniew, Jack Lindsey and co-authors; Anthropic",
      "publishedAt": "2026-07-06",
      "version": "Research paper: 6 July 2026; arXiv:2607.15495v1: 16 July 2026",
      "type": "Preprint",
      "reviewLevel": "Selected sections",
      "finding": "The authors identify a reportable subspace, J-space, involved in unspoken intermediate computation and flexible reasoning. Interventions alter some decisions, while some routine processing continues outside this subspace.",
      "limitations": "A functional resemblance to a global workspace does not establish subjective experience. The lens has limits in reading concepts and defining layer boundaries. Reportable representations need not already be written sentences.",
      "changes": "The work separates visible output from internal computation. It does not demonstrate a universal language independent of all human concepts.",
      "checkedAt": "2026-09-30",
      "verificationStatus": "Primary source reviewed",
      "relatedUrls": [
        "https://arxiv.org/abs/2607.15495"
      ]
    },
    {
      "id": "R-AG-02",
      "title": "Enabling Agents to Communicate Entirely in Latent Space",
      "url": "https://aclanthology.org/2026.acl-long.1248/",
      "authors": "Zhuoyun Du, Runze Wang, Huiyu Bai, Zouying Cao, Xiaoyong Zhu, Yu Cheng, Bo Zheng, Wei Chen, and Haochao Ying",
      "publishedAt": "2026-07",
      "version": "ACL publication metadata: July 2026; inspected experimental text: arXiv:2511.09149v2, 7 January 2026",
      "type": "Peer-reviewed conference paper; experimental review uses an earlier preprint",
      "reviewLevel": "Selected sections",
      "finding": "Interlat v2 reports trained hidden-state communication, including a Qwen-to-LLaMA experiment. MATH Table 2 reports 36.88% overall for Interlat versus 38.35% for full CoT, while Level 5 is 15.80% versus 15.05%.",
      "limitations": "Internal model access, adapters and training are required. The results do not establish arbitrary-model interoperability or preservation of ethical distinctions. The final ACL PDF was inaccessible; the inspected v2 is not assumed identical to the conference paper or later revisions.",
      "changes": "Rereview on 4 October extends the 30 September note: v2 setup, Tables 1–2 and training discussion were read, plus ACL metadata. Final ACL PDF was inaccessible; later preprints were not reviewed.",
      "checkedAt": "2026-10-04",
      "verificationStatus": "Primary preprint sections and conference metadata reviewed; final PDF inaccessible; no reproduction",
      "relatedUrls": [
        "https://arxiv.org/html/2511.09149v2",
        "https://arxiv.org/abs/2511.09149v2",
        "https://doi.org/10.18653/v1/2026.acl-long.1248"
      ]
    },
    {
      "id": "R-AG-03",
      "title": "Latent Collaboration in Multi-Agent Systems",
      "url": "https://arxiv.org/html/2511.20639v4",
      "authors": "Jiaru Zou and co-authors",
      "publishedAt": "2025-11-25",
      "version": "LatentMAS; arXiv v4: 3 August 2026",
      "type": "Preprint",
      "reviewLevel": "Selected sections",
      "finding": "The study examines collaboration through hidden-state and key-value-cache transfer without additional training. It reports lower token use and latency, with accuracy improvements on several reasoning benchmarks.",
      "limitations": "Appendix G assumes the same transformer architecture; adapters across different models are future work. Expressiveness claims depend on theoretical assumptions, and lossless transfer is not a guarantee of correctness.",
      "changes": "The first publication was November 2025; the reviewed revision is from August 2026. Direct use between arbitrary closed model APIs is not demonstrated.",
      "checkedAt": "2026-09-30",
      "verificationStatus": "Primary source reviewed",
      "relatedUrls": [
        "https://arxiv.org/abs/2511.20639",
        "https://github.com/Gen-Verse/LatentMAS"
      ]
    },
    {
      "id": "R-AG-01",
      "title": "Zerolang",
      "url": "https://github.com/vercel-labs/zerolang",
      "authors": "Vercel Labs",
      "publishedAt": null,
      "version": "Repository README checked 30 September 2026; no commit pinned",
      "type": "Software prototype",
      "reviewLevel": "Project README",
      "finding": "This agent-oriented language prototype makes a semantic graph or database the authoritative program, with readable views and operations for structural editing.",
      "limitations": "The prototype is a design example, not evidence of broad performance gains or universal language convergence. Its code was not run, and its initial release date was not verified.",
      "changes": "It illustrates how readable source text can become one view of a program rather than its only editing surface.",
      "checkedAt": "2026-09-30",
      "verificationStatus": "Primary source reviewed",
      "relatedUrls": []
    },
    {
      "id": "R-GOV-06",
      "title": "Models as Insider Risks in the Super Intelligence Era",
      "url": "https://snscratchpad.com/posts/models-as-insider-risks/",
      "authors": "Satya Nadella",
      "publishedAt": "2026-10-10",
      "version": "Author blog essay dated 10 October 2026; read 11 October 2026; no numbered revision",
      "type": "Architectural opinion essay",
      "reviewLevel": "Full essay",
      "finding": "Nadella argues for separating intelligence from authority, with external controls, independent action evidence and audits, containment, model diversity and incident disclosure.",
      "limitations": "An architectural argument, not an experiment or validation of Leviathan. Controls need evaluation; neither model diversity nor readable reasoning guarantees independent or faithful evidence. The essay does not evaluate local meaning kernels or governance among independent communities.",
      "changes": "Read the full author-hosted essay. The supplied X link was inaccessible to the reading tool; the blog was retrieved directly. Publication and review dates remain separate.",
      "checkedAt": "2026-10-11",
      "verificationStatus": "Primary source reviewed; architectural proposals not independently validated",
      "relatedUrls": [
        "https://x.com/satyanadella/status/2108931348857827686"
      ]
    },
    {
      "id": "R-AG-10",
      "title": "LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence-Conclusion Coevolution",
      "url": "https://academy.dair.ai/papers/longharness-advancing-agentic-reasoning-over-100m-token-corpora-via-stateful-evi-2610.13075",
      "authors": "Shanghai AI Laboratory, Fudan University, Zhejiang University, and Shanghai Jiao Tong University researchers (as indexed by DAIR.AI)",
      "publishedAt": "2026-10",
      "version": "arXiv:2610.13075; indexed in DAIR.AI Papers of the Week (Oct 12–Oct 18, 2026)",
      "type": "Preprint",
      "reviewLevel": "Indexed technical digest (DAIR.AI Papers of the Week 2026)",
      "finding": "LongHarness separates retrieved evidence and evolving conclusions into explicit persistent state rather than leaving them mixed in interaction history. The authors identify evidence-state drift—where earlier evidence or later contradicting discoveries fail to update current conclusions—and evaluate Evidence-Conclusion Coevolution on corpora up to 100M tokens.",
      "limitations": "This review is based on the curated DAIR.AI technical summary rather than a full-text audit or benchmark reproduction. Stateful evidence tracking for single-agent document synthesis does not by itself validate Leviathan’s multi-participant meaning kernels or governance.",
      "changes": "Motivates separating explicit evidence state and contradiction propagation from raw conversation history when conclusions are revised over time.",
      "checkedAt": "2026-10-21",
      "verificationStatus": "Indexed technical summary reviewed (DAIR.AI 2026); primary full text and experiments not independently reproduced",
      "relatedUrls": [
        "https://github.com/dair-ai/AI-Papers-of-the-Week/blob/main/years/2026.md"
      ]
    },
    {
      "id": "R-AG-14",
      "title": "Selective Forgetting: A Graph-Based Memory Framework for Long-Term LLM Agents",
      "url": "https://academy.dair.ai/papers/selective-forgetting-a-graph-based-memory-framework-for-long-term-llm-agents-2608.28978",
      "authors": "Authors as indexed in DAIR.AI Papers of the Week (arXiv:2608.28978)",
      "publishedAt": "2026-08",
      "version": "arXiv:2608.28978; indexed in DAIR.AI Papers of the Week (Aug 31–Sep 6, 2026)",
      "type": "Preprint",
      "reviewLevel": "Indexed technical digest (DAIR.AI Papers of the Week 2026)",
      "finding": "Under a fixed 4,096-token retrieval budget on LongMemEval, a decaying knowledge-graph memory underperforms a flat vector baseline (0.42 vs. 0.47 overall F1, and 0.607 vs. 0.911 judged correctness when answers depend on prior assistant turns) because entity extraction discards surface phrasing and multi-entity context.",
      "limitations": "Reviewed via the curated DAIR.AI technical summary; experiments were not reproduced. The negative result applies to a specific LLM-extracted entity graph and decay schedule on conversational memory benchmarks, not to all structured provenance or versioned claim records.",
      "changes": "Challenges assuming that decomposing text into a graph automatically improves retrieval or reasoning over simpler text-preserving baselines.",
      "checkedAt": "2026-10-21",
      "verificationStatus": "Indexed technical summary reviewed (DAIR.AI 2026); primary full text and experiments not independently reproduced",
      "relatedUrls": [
        "https://github.com/dair-ai/AI-Papers-of-the-Week/blob/main/years/2026.md"
      ]
    }
  ]
}
