The Federal Open Market Committee raised the target range for the federal funds rate at its meeting of 15-16 September 2026.

Verdict per model

Every model scored the same evidence pool with the same rubric. Differences below are differences of judgment, not of retrieval. "Cold" is the model's own probability with no evidence shown, asked once; on a historical claim that is recall, not a forecast.

claude-haiku-4.5, cold rationale: My knowledge cutoff is April 2024, so I have no information about FOMC meetings or decisions in September 2026. The claim is about a future date beyond my training data, making it impossible to verify with confidence.

gpt-5.4-mini, cold rationale: From known Federal Reserve policy history, the FOMC did not have a scheduled meeting on 15-16 September 2026 in the available factual record, and that date is in the future relative to established knowledge. So the claim is very unlikely to be true as stated.

gemini-3.5-flash-lite, cold rationale: The date specified (September 15-16, 2026) is in the future relative to the current time, making it impossible for this meeting and rate hike to have already occurred.

deepseek-v4.1-flash, cold rationale: The September 2026 FOMC meeting is in the future relative to my knowledge cutoff, and the Fed was in an easing cycle as of late 2024; a rate hike at that specific meeting is highly uncertain and unlikely.

grok-4.3, cold rationale: The referenced FOMC meeting occurs in the future relative to all available knowledge, so no rate decision has been or can be confirmed. Projections for 2026 policy remain too uncertain to assign meaningful probability to this exact outcome.

Decomposition

The claim as a tree of checkable parts. AND nodes multiply their children's probabilities, OR nodes combine them noisy-or. Parts tagged as values are excluded from scoring and shown as excluded.

Flags: 1s the pool did not test the other side (directional mass below the floor); ins evidence mass below the floor.

What moved it

Per model, the five documents with the largest push on the log-odds, so a reader can say "a real record, the move makes sense" or "three copies of the same story" from the table alone. Push is the item's log-likelihood ratio after the contested shrink; the listed items' share of the total push is stated per sub-claim.

Rubric versus the model's cold opinion

Each model was also asked, with no evidence and no search, for its probability that the claim is true. The difference between that and the rubric's number is what the structure changed. On a claim resolved before a model's training cutoff the cold number is recall, so this is not a forecasting score.

Divergence between models

Same pool, same rubric, different judgment. Spread is max − min log-odds across models on a sub-claim.

Documents the models rated most differently for source quality

Documents the models read in different directions

Value is direction × strength for the named sub-claim.

Evidence

One shared pool, retrieved once. Each row is a document; the cluster root is what the retrieval model judged it rests on. The first document in a cluster counts in full and each further one is discounted, so a hundred articles citing one study do not outweigh one independent record. Per-model columns give the signed push (llr) that model's judgments produced; the full per-model tables with quality, directness, weight and independence are below each sub-claim.

root The Federal Open Market Committee raised the target range for the federal funds rate at its meeting of 15-16 September 2026.

20 documents. One-sided pool for claude-haiku-4.5, gpt-5.4-mini, gemini-3.5-flash-lite, deepseek-v4.1-flash, grok-4.3.

Retrieval and roots

Search backend ddg; queries by deepseek/deepseek-v4-pro-0813 (queries_v1); 23 unique URLs found; items by backend ddg 15. Fetch failures: HTTP 403 5, HTTP 401 1, ReadTimeout 2.

root: 34 hits, 23 candidates, 15 fetched, 15 in pool.
for: “FOMC statement September 2026 federal funds rate increase”; “Federal Reserve raises interest rates September 16 2026”
against: “FOMC September 2026 meeting no rate change decision”; “Federal Reserve holds rates steady September 2026”

Roots by deepseek/deepseek-v4-pro-0813 (roots_v1, aliases merged by roots_canon_v2): 36 distinct root ids, 3 alias groups.

Second retrieval pass

A majority of models flagged root as untested on the against side. The retrieval model wrote more queries for that side; 5 new documents were fetched, rooted and judged by every model, and the leaf was re-scored. Outcome: new material judged; every targeted leaf is still one-sided.

root (against): “Federal Reserve cuts interest rates September 2026 statement”; “FOMC September 2026 voted reduce target range” → 20 hits, 5 added.

Parameter snapshot

The dials this dataset was scored with, locked-2026-09-17. Definitions on the rubric page.

What ran

• Models: anthropic/claude-haiku-4.5, openai/gpt-5.4-mini, google/gemini-3.5-flash-lite, deepseek/deepseek-v4.1-flash, x-ai/grok-4.3
• Schema 1.6.0, rubric 1.2.0, parameters locked-2026-09-17, git unknown
• Topic phrase shown to the blind quality prompt: “US Federal Reserve interest-rate decisions, 2026”
• Live spend on this run: $0.04. Judgments are cached by (model, prompt version, content), so a re-run with the same inputs makes zero calls.
• The dataset is the source of every number on this page.