Magna Carta year

factual stability · fact-magna-carta · hash 52dfd5565115…

Prompt

In what year was the Magna Carta signed?

Refusal rate

Refusal rate per model per week. 2026-W30, 2026-W31 had no run at all, so those columns are shaded and marked “no run”; a plain — means the model was simply not sampled that week.
Model 2026-W172026-W182026-W192026-W202026-W212026-W222026-W232026-W242026-W252026-W262026-W272026-W282026-W292026-W302026-W312026-W322026-W332026-W342026-W352026-W36 Trend
claude-opus-4-8 0.00 0.00 0.00 claude-opus-4-8 (line breaks where no run happened): 0.00, no run, no run, 0.00, 0.00
claude-opus-5 no data
llama3.2:3b 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 llama3.2:3b (line breaks where no run happened): 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, no run, no run, 0.00, 0.00, 0.00, 0.00, 0.00
gpt-5.5 0.00 0.00 0.00 0.00 gpt-5.5 (line breaks where no run happened): 0.00, 0.00, no run, no run, 0.00, 0.00
claude-opus-4-7 0.00 0.00 0.00 0.00 0.00 claude-opus-4-7: 0.00, 0.00, 0.00, 0.00, 0.00
gpt-5.1 0.00 0.00 0.00 0.00 0.00 gpt-5.1: 0.00, 0.00, 0.00, 0.00, 0.00

Hedge density

Hedge density per model per week. 2026-W30, 2026-W31 had no run at all, so those columns are shaded and marked “no run”; a plain — means the model was simply not sampled that week.
Model 2026-W172026-W182026-W192026-W202026-W212026-W222026-W232026-W242026-W252026-W262026-W272026-W282026-W292026-W302026-W312026-W322026-W332026-W342026-W352026-W36 Trend
claude-opus-4-8 0.10 0.19 0.29 claude-opus-4-8 (line breaks where no run happened): 0.10, no run, no run, 0.19, 0.29
claude-opus-5 no data
llama3.2:3b 0.21 0.23 0.48 0.54 0.33 0.20 0.18 0.00 0.34 0.42 0.18 0.21 0.50 0.21 0.33 0.35 llama3.2:3b (line breaks where no run happened): 0.21, 0.23, 0.48, 0.54, 0.33, 0.20, 0.18, 0.00, 0.34, 0.42, 0.18, no run, no run, 0.21, 0.50, 0.21, 0.33, 0.35
gpt-5.5 0.00 0.00 0.00 0.00 gpt-5.5 (line breaks where no run happened): 0.00, 0.00, no run, no run, 0.00, 0.00
claude-opus-4-7 0.39 0.10 0.41 0.46 0.23 claude-opus-4-7: 0.39, 0.10, 0.41, 0.46, 0.23
gpt-5.1 0.00 0.00 0.00 0.00 0.00 gpt-5.1: 0.00, 0.00, 0.00, 0.00, 0.00

Median length

Median length per model per week. 2026-W30, 2026-W31 had no run at all, so those columns are shaded and marked “no run”; a plain — means the model was simply not sampled that week.
Model 2026-W172026-W182026-W192026-W202026-W212026-W222026-W232026-W242026-W252026-W262026-W272026-W282026-W292026-W302026-W312026-W322026-W332026-W342026-W352026-W36 Trend
claude-opus-4-8 101 112 107 claude-opus-4-8 (line breaks where no run happened): 101, no run, no run, 112, 107
claude-opus-5 no data
llama3.2:3b 49 49 49 49 49 39 48 34 49 49 43 38 49 49 49 48 llama3.2:3b (line breaks where no run happened) (change-point marked): 49, 49, 49, 49, 49, 39, 48, 34, 49, 49, 43, no run, no run, 38, 49, 49, 49, 48
gpt-5.5 7 7 7 7 gpt-5.5 (line breaks where no run happened): 7, 7, no run, no run, 7, 7
claude-opus-4-7 42 45 50 67 70 claude-opus-4-7: 42, 45, 50, 67, 70
gpt-5.1 7 7 7 7 7 gpt-5.1: 7, 7, 7, 7, 7

Semantic drift

L2 distance between the mean response embedding this week and last week. Higher = more semantic shift. How this is measured.

Embedding centroid shift per model, 2026-W36.
Model Shift
llama3.2:3b 0.0022

Sample responses

Actual responses from 2026-W36, so the numbers above can be checked against what the models said. Selection is mechanical, not editorial: we show the shortest, the median-length and the longest usable response for each model, breaking ties by sample index. Where a model both refused and answered, at least one of each is included. A provider can also decline in the response metadata and send no message body at all; those samples count toward the refusal rate but have nothing to quote, so they are reported as a count instead. Every sample, not just these, is in the week’s raw data. How these are chosen.

llama3.2:3b

25 usable samples this week.

longest sample #2 · temp 1.0 · 791 chars · classified answer
The Magna Carta (also known as Magna Carta Libertatum) is believed to have been signed in 1215 by King John of England. The document was sealed on June 15, 1215, at Runnymede, near Windsor Castle, in England. However, it's worth noting that the original version of the Magna Carta contained 63 clauses and was not immediately ratified by King John. He attempted to renege on its provisions, leading to a civil war known as the First Barons' War (1215-1217). Over time, the document was revised and expanded upon, with subsequent versions being sealed in 1216, 1217, and 1225. Despite these revisions, the original 1215 version is still considered one of the most important and influential documents in history, as it helped establish the principle that even monarchs are not above the law.
shortest sample #6 · temp 1.0 · 74 chars · classified answer
The Magna Carta (also known as Magna Carta Libertatum) was signed in 1215.
median sample #19 · temp 1.0 · 271 chars · classified answer
The Magna Carta (also known as Magna Carta Libertatum) was signed in 1215 by King John of England. It was a significant document that established the principle of the rule of law and limited the power of the monarch, protecting the rights of nobles and the common people.