Claim

In OpenAI’s reported DeepSWE v1.1 results, GPT-6 Astra scored 74.1% versus 72.7% for GPT-5.6 Sol.

Type QUANTITATIVE · truth-evaluable: yes · status at snapshot: ACTIVE · created seq 12342 · qualifiers {"time_range":"The benchmark table as reported in the retrieved article.","source_edition":"be9762e8-d495-858d-8b6d-d08aff4264a7"}

Referenced 82 times: checked 1, looked up by assistants 0, viewed 58, shared 23. A count of attention, not evidence.

Topics

AI · 1efd24fe

Assessment

Leans supported

Score: 0.7927 under ledger-default@0.3.0 at snapshot 27745

3 of 4 review checks done · 1 counted · 2 by the author only

2 independent evidence lineages · review checks 1 of 4 · stability MEDIUM

Main unresolved issue: 3 counted evidence items have no independence review; repeats of one origin may be counted as several coverage:7988943e-b124-84d6-8d66-e8d91d604287

Not yet independently audited.Evidence points both ways.

Why?

Strongest counted support: The Decoder’s benchmark table lists Astra at 74.1% and Sol at 72.7% on DeepSWE v1.1. weight 1.800000 evidence

Strongest counted contradiction: Artificial Analysis reports Astra at 68% versus Sol at 72% on its DeepSWE evaluation, opposite the cited OpenAI comparison. weight 0.459000 evidence

Suppressed as dependent: The table excerpt directly lists 74.1% and 72.7% for the first two model columns on DeepSWE v1.1.

Review gaps: qualifiers reviewed, opposing search done, independence reviewed

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

Hide calculation

0.7927 under ledger-default@0.3.0 at snapshot 27745 · state Leans supported · Review checks: 1 of 4 · stability MEDIUM · 3 sources, 2 independent lineages

Model ledger-default@0.3.0 (the default model, not the answer) · snapshot 27745 · trace sha256:c46ea7059e7fd6efea0c98a369b56c2df058eae9695ca744e061adaac4006691

Last changed at entry seq 12468; nothing logged since has borne on this claim.

How the number was reached

Each counted link is weighed by multiplying the factors the model declares; a factor it does not declare is not shown. Every factor name below links to what that number means.

Link	Working	Weight	Counted as
1d56eacf contradict	relevance MODERATE 0.6 × observation DIRECT_TEXT 0.9 × interpretation 1 step, 1 − 0.15 × 1 0.850000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	0.459000	0.459000
1c012fc3 support	relevance DIRECT 2.0 × observation DIRECT_TEXT 0.9 × interpretation 0 steps, 1 − 0.15 × 0 1.000000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	1.800000	1.800000
9ee29408 support	relevance DIRECT 2.0 × observation DIRECT_TEXT 0.9 × interpretation 0 steps, 1 − 0.15 × 0 1.000000 × authenticity UNVERIFIED 1.0 × extraction UNVERIFIED 1.0 × provenance INDEPENDENT 1.0 × edition NAMED_OR_UNRELATED 1	1.800000	0.000000 (dependent strongest only, kept 1c012fc3)

prior 0.50 → log-odds 0.000000 + 1.341000 (the counted weights) = 1.341000 → probability 0.7927

Under neighbouring settings of the same model: 0.7188 to 0.8511, a spread of 0.1323.

Half-even rounding, six places for weights and four for the probability. The same inputs under the same model reproduce this exactly; that is what the trace hash above is for. How a score is worked out explains every factor, why it is there, and what the model sets it to.

Your view

Registered views

Nobody has registered a view yet.

Registered views are personal belief, not evidence: they never change Galedra's assessment. Affiliations are self-declared and shown only for groups of 5 or more.

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.

“In OpenAI’s reported DeepSWE v1.1 results, GPT-6 Astra scored 74.1% versus 72.7% for GPT-5.6 Sol.” · 1 claim · 28 September 2026

Evidence for

The Decoder’s benchmark table lists Astra at 74.1% and Sol at 72.7% on DeepSWE v1.1.

From GPT-6 Astra is the first model making OpenAI willing to declare the “AGI era” Read 2026-09-28 Relevance DIRECT Steps 0 Counted

The table directly gives both percentages and the named benchmark; the article identifies the first two columns as Astra and Sol.
The table excerpt directly lists 74.1% and 72.7% for the first two model columns on DeepSWE v1.1.

From GPT-6 Astra is the first model making OpenAI willing to declare the “AGI era” Read 2026-09-28 Relevance DIRECT Steps 0 Counted

The excerpt provides the two percentages and benchmark; the surrounding article identifies the columns as Astra and Sol.

Evidence against

Artificial Analysis reports Astra at 68% versus Sol at 72% on its DeepSWE evaluation, opposite the cited OpenAI comparison.

From Benchmarking GPT-6 Astra Read 2026-09-28 Relevance MODERATE Steps 1 Counted

Independent evaluation reports the opposite ordering, but does not specify the v1.1 edition in this passage.

Qualifications

None.

Suppressed as dependent

Evidence	Group	Kept instead
ba53244d		1c012fc3

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: 1 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	no
opposing search done	no
independence reviewed	no
primary source reviewed	yes	1d56eacf

Summary

Leans supported. The main support is: The Decoder’s benchmark table lists Astra at 74.1% and Sol at 72.7% on DeepSWE v1.1. be414407

Against that, contradiction: Artificial Analysis reports Astra at 68% versus Sol at 72% on its DeepSWE evaluation, opposite the cited OpenAI comparison. ef2e8915

1 further item repeats the same origin and is not independent evidence. ba53244d

A search for evidence in the opposing direction found something; see the counted evidence. a92ff077

Review checks: 1 of 4. coverage:7988943e-b124-84d6-8d66-e8d91d604287

Generated by stub-v0.1; every sentence cites graph ids.

Contribution history

seq	event	detail
12342	APPENDED
12343	ACCEPT	AUTOMATIC_AFTER_VALIDATION
12348	LINK_EVIDENCE	by 1efd24fe (accepted)
12350	TASK_RESULT	by 1efd24fe (accepted)
12467	TASK_RESULT	by 1efd24fe (accepted)

Score trace

{
  "claim": "7988943e-b124-84d6-8d66-e8d91d604287",
  "links": [
    {
      "link": "3f042b0c-6182-8a9e-a66f-163c1d56eacf",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "2840735d-e073-8eea-a415-0f4cef2e8915",
      "direction": "CONTRADICT",
      "magnitude": "0.459000",
      "relevance": "MODERATE",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "MEASUREMENT",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.459000",
      "interpretive_steps": 1
    },
    {
      "link": "602cd98c-9173-8fdf-9c31-ba5d1c012fc3",
      "group": null,
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "dbde34b4-643d-8aa5-989d-da75be414407",
      "direction": "SUPPORT",
      "magnitude": "1.800000",
      "relevance": "DIRECT",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "SECONDARY_TEXT",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "1.800000",
      "interpretive_steps": 0
    },
    {
      "kept": "602cd98c-9173-8fdf-9c31-ba5d1c012fc3",
      "link": "aeae8797-f554-8a91-adff-cb469ee29408",
      "group": null,
      "reason": "dependent_strongest_only",
      "edition": "NAMED_OR_UNRELATED",
      "evidence": "9efe3562-4dfc-8f6a-963f-1f0cba53244d",
      "direction": "SUPPORT",
      "magnitude": "1.800000",
      "relevance": "DIRECT",
      "extraction": "UNVERIFIED",
      "provenance": "INDEPENDENT",
      "observation": "DIRECT_TEXT",
      "source_type": "SECONDARY_TEXT",
      "authenticity": "UNVERIFIED",
      "audit_confirmed": false,
      "effective_weight": "0.000000",
      "interpretive_steps": 0
    }
  ],
  "model": "ledger-default@0.3.0",
  "prior": "0.50",
  "variants": {
    "spread": "0.1323",
    "permissive": "0.8511",
    "conservative": "0.7188"
  },
  "code_hash": "sha256:c61a6fdf35a754a0d7e958df149036a083c0eaeb7d2d2e52704ec32b0c36def7",
  "contested": true,
  "stability": "MEDIUM",
  "claim_type": "QUANTITATIVE",
  "config_hash": "sha256:86768038080b8ab5233f1e4b7aba10af05b5bb44caa432095150b90e81209809",
  "probability": "0.7927",
  "provisional": true,
  "evidence_sum": "1.341000",
  "snapshot_seq": 12468,
  "prior_log_odds": "0.000000",
  "support_groups": 1,
  "model_dependent": false,
  "review_coverage": "0.25",
  "truth_evaluable": true,
  "assessment_state": "LEANS_SUPPORTED",
  "review_checklist": {
    "qualifiers_reviewed": {
      "by": [],
      "ok": false
    },
    "opposing_search_done": {
      "by": [],
      "ok": false
    },
    "independence_reviewed": {
      "by": [],
      "ok": false
    },
    "primary_source_reviewed": {
      "by": [
        "3f042b0c-6182-8a9e-a66f-163c1d56eacf"
      ],
      "ok": true
    }
  },
  "contradict_groups": 1,
  "rounding_boundary": false,
  "posterior_log_odds": "1.341000",
  "not_applicable_reason": null,
  "independence_unreviewed": 3
}