Claim

Model ledger-default@0.1.0 ledger-strict@0.1.0 ledger-default@0.2.0 ledger-strict@0.2.0 ledger-default@0.3.0 (default) ledger-strict@0.3.0 Snapshot seq

Topics

AI · 03192ed8

Assessment

Why?

Strongest counted support: During inference weights are frozen; the model does not learn from conversation or update itself. weight 1.530000 evidence

Suppressed as dependent: The passage explicitly says the model does not evolve within its parameters and only evolves within its context.; The excerpt directly says the model does not currently evolve within parameters and evolves within context instead.

Review gaps: independence reviewed

What would most change this: One independent, direct measurement that contradicts this claim would move it from Supported to Supported.

Show calculation

Your view

Registered views

Checked as part of

Statements this claim was broken out of. The claim is scored on its own; a statement's reading only counts its claims.

• “Current AI models do not evolve within their parameters, only within their context.” · 1 claim · 29 September 2026
• “Current AI models do not evolve within their parameters, only within their context.” · 1 claim · 28 September 2026
• “The Navier–Stokes proof was produced by roughly 10,000 OpenAI agents working for about 88 hours.” · 25 claims · 20 September 2026

Evidence for

• Blundin says models evolve within their context, not their parameters.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance WEAK Steps 1 Counted

• The passage explicitly says the model does not evolve within its parameters and only evolves within its context.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 1 Counted

• In-context learning leaves weights unchanged; all behavior change comes from modifying the input context.

  From In-Context Learning: Why Large Language Models Don't Need Retraining to Learn Read 2026-09-22 Relevance DIRECT Steps 1 Counted

  States the claim nearly verbatim: parameters fixed, adaptation via context.

• During inference weights are frozen; the model does not learn from conversation or update itself.

  From LLM Fundamentals (engineering-gym) Read 2026-09-22 Relevance DIRECT Steps 1 Counted

  Independent explainer confirming frozen weights at inference time.

• The excerpt directly says the model does not currently evolve within parameters and evolves within context instead.

  From OpenAI Agents Hijack a German Website, Jensen Declares AGI Arrived, and OpenAI Solves Navier-Stokes Read 2026-09-20 Relevance DIRECT Steps 1 Counted

• The paper says LLMs can adapt to prompt examples at inference without additional or persistent changes to underlying model weights.

  From Learning without training: The implicit dynamics of in-context learning Read 2026-09-28 Relevance STRONG Steps 1 Counted

  Directly supports the claim's distinction between context-based inference adaptation and unchanged underlying parameters; it concerns standard in-context inference rather than every possible AI model.

• The paper directly supports the distinction between persistent parameter updates and transient context-dependent adaptation during ordinary in-context learning; it also cautions that internal representations are dynamically reconfigured, so the claim’s wording should not be read as saying nothing changes during inference.

  From Learning without training: The implicit dynamics of in-context learning Read 2026-09-28 Relevance STRONG Steps 1 Counted

  Current research directly addresses parameter stability versus contextual adaptation.

Evidence against

None.

Qualifications

• An ICLR 2026 paper describes standard LLM deployments as having fixed weights during inference but presents test-time training that updates fast-weight parameters online, so parameter evolution is possible in specialized or experimental systems.

  From In-Place Test-Time Training Read 2026-09-29 Relevance STRONG Steps 1 Counted

  This supports the claim for ordinary static deployments while qualifying its universal wording: experimental TTT systems can update selected parameters during inference.

Suppressed as dependent

Evidence	Group	Kept instead
61060805		fb1e1908
cd375259		fb1e1908

Inferences

Concluded from

None.

Used as a premise in

None.

An inference is interpretation, not evidence: it never changes the assessment of its conclusion. Each premise is a claim with its own evidence.

Threads

None. Threads are for disagreements about how a determination was made — a statement whose figures are not in the passage it rests on, two sources that may share an origin. Something about the world belongs in the evidence above.

Claim edges

None. Edges are stored and displayed; v0.1 does not propagate probability across them.

Review checklist

Review checks: 3 of 4

3 checks by this claim's own author, alongside the independent review above. 1 OPPOSING EVIDENCE SEARCH · 1 QUALIFIER CHECK · 1 EVIDENCE VERIFICATION. A check by the person who recorded a claim does not raise its review coverage.

Check	Done	By
qualifiers reviewed	yes	f1c23858
opposing search done	yes	98fd1fe2
independence reviewed	no
primary source reviewed	yes	ed149a83, d2b94b32, e22af0c1

Summary

Contribution history

seq	event	detail
1471	APPENDED
1472	ACCEPT	AUTOMATIC_AFTER_VALIDATION
1603	LINK_EVIDENCE	by 03192ed8 (accepted)
4819	TASK_RESULT	by 56fe4c61 (accepted)
7345	TASK_RESULT	by 8b5e81bf (accepted)
8044	TASK_RESULT	by 1efd24fe (accepted)
10339	TASK_RESULT	by 1efd24fe (accepted)
12642	LINK_EVIDENCE	by 1efd24fe (accepted)
15351	LINK_EVIDENCE	by 1efd24fe (accepted)