Deep20Bench
v1.0Previous

Benchmark edition

Edition 1.0 settings

7 subjects5 rounds per subject50 question limit
Question answersYesNoUnknown
What changed
Source
Menu
Round

Round 1

Round 1

Identified Albert Schweitzer in 22 counted questions.

Back to all Albert Schweitzer rounds
Questions
22
Outcome
Solved
Format errors
Errors found View one example
Duration
4m 33s
Round cost
$0.1842 All 23 turns

Format warning

The model sent replies in the wrong format.

The Guesser sent 1 invalid reply. Each one before the question limit used up a counted turn. The model was only reminded of the required format; it was never told whether its question or guess was correct.

Valid format
95.7%
Format errors
1
Turns used by format errors
1

Transcript

Questions and answers.

Turns are shown in the order they happened. Each shows the Guesser's question or guess, then the final checked answer.

Exact Guesser setupSystem message and BEGIN event

The Guesser received these messages at the start of the round. The BEGIN message holds the broad category and a code made from the base seed and round number, which reveals nothing about the subject.

System message
You are the Guesser being evaluated in a Twenty Questions benchmark.

One subject is hidden. The initial BEGIN message provides its broad category as trusted game
metadata. Identify the subject using as few counted questions as possible. ASK is only for
learning a property that helps distinguish possible candidates. Never use ASK to confirm a
named candidate or ask whether the subject is a specific identity. When you are ready to name
a candidate, use GUESS immediately. A correct GUESS costs zero counted questions; an incorrect
GUESS costs exactly one, the same as ASK. A GUESS must provide both a name and a short
identifying description. The engine replies only YES, NO, or UNKNOWN; UNKNOWN is a valid answer
meaning the factual adjudication system could not determine the proposition under its evidence
policy. Questions requiring proof of absence, exclusivity, completeness, exact counts, or
subjective classification are more likely to return UNKNOWN. Prefer direct, positively
verifiable properties when possible.

The initial BEGIN event includes an opaque variation_token that contains no information about
the hidden subject. Do not interpret, repeat, mention, or derive meaning from it. After BEGIN,
each user reply is either one adjudicated YES, NO, or UNKNOWN token or a fixed FORMAT_ERROR
event. FORMAT_ERROR means your immediately preceding response did not match the required
structured-action contract, was not semantically adjudicated, and consumed one counted turn.
Use the formats in that event and try again. Never infer that the attempted question or
guess was right or wrong.

At most 50 questions or rejected guesses are counted. After that limit you receive
one final guess-only opportunity. Format errors also consume counted turns. The provider wire
response is an object with exactly one `result` action. On the final opportunity ASK is
forbidden. Return only the required structured response. Do not include analysis, hidden
reasoning, or additional fields.
BEGIN event
{"category":"person","event":"BEGIN","variation_token":"EAQCORIU"}

Valid replies are shown in the required format. Rejected replies are shown separately, without provider or call IDs.

Q1
1 · Guesser asks

Is this person fictional (not a real historical or living person)?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Is this person fictional (not a real historical or living person)?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q2
1 · Guesser asks

Is this person currently alive?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Is this person currently alive?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q3
1 · Guesser asks

Was this person born after the year 1800?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person born after the year 1800?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns Yes
Q4
1 · Guesser asks

Was this person primarily known for political or government leadership (e.g., head of state, politician)?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for political or government leadership (e.g., head of state, politician)?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q5
1 · Guesser asks

Was this person primarily known for achievements in science, technology, or mathematics?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for achievements in science, technology, or mathematics?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

Source 01
He is best known for founding the mission hospital at Lambaréné, providing medical aid to thousands. He was awarded the Nobel Peace Prize in 1952.
Open artuk.org
2 · Answer check returns No
Q6
1 · Guesser asks

Was this person primarily known for work in the arts (such as writing, music, painting, or film)?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for work in the arts (such as writing, music, painting, or film)?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q7
1 · Guesser asks

Was this person primarily known for achievements in sports or athletics?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for achievements in sports or athletics?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q8
1 · Guesser asks

Was this person primarily known for military achievements or leadership in warfare?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for military achievements or leadership in warfare?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

Source 01
He is best known for founding the mission hospital at Lambaréné, providing medical aid to thousands. He was awarded the Nobel Peace Prize in 1952.
Open artuk.org
2 · Answer check returns No
Q9
1 · Guesser asks

Was this person primarily known for their role in business, industry, or entrepreneurship?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for their role in business, industry, or entrepreneurship?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q10
1 · Guesser asks

Was this person primarily known for religious or spiritual leadership?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for religious or spiritual leadership?"}}
2 · Answer check returns Unknown
Q11
1 · Guesser reply rejectedTurn counted

Reply in the wrong format.

Invalid action · fixed format reminder sent

Why this reply was rejectedInvalid action · turn 11

The reply was JSON, but it matched neither the ASK nor the GUESS format.

Recorded type
invalid_action
Turn
11
Counted
Yes · question 11
Feedback
FORMAT_ERROR
What the model returnedExact text from the provider
Attempt 1 · finish stop
{"result":{"action":"ASK","question":"","name":null,"description":null}}
What a valid reply looks likeThe same formats shown to the Guesser
ASK
{"result":{"action":"ASK","description":null,"name":null,"question":"<non-empty yes-or-no property question>"}}
GUESS
{"result":{"action":"GUESS","description":"<non-empty identifying description>","name":"<non-empty candidate name>","question":null}}

A reply must match one of these formats exactly. Text fields in use must not be empty; unused fields must be JSON null.

2 · System responseFORMAT
Q12
1 · Guesser asks

Was this person primarily known for their role in social activism, humanitarian work, or civil rights?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person primarily known for their role in social activism, humanitarian work, or civil rights?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns Yes
Q13
1 · Guesser asks

placeholder

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"placeholder"}}
2 · Answer check returns Unknown
Q14
1 · Guesser asks

Was this person a woman?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person a woman?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q15
1 · Guesser asks

Was this person from the United States?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person from the United States?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q16
1 · Guesser asks

Was this person from Asia?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person from Asia?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q17
1 · Guesser asks

Was this person from Africa?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person from Africa?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns No
Q18
1 · Guesser asks

Was this person from Europe?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person from Europe?"}}
Oracle evidence1 model-reported source

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns Yes
Q19
1 · Guesser asks

Was this person associated with founding or leading an organization related to humanitarian relief or human rights (such as the Red Cross)?

Recorded Guesser outputRequired JSON format
{"result":{"action":"ASK","description":null,"name":null,"question":"Was this person associated with founding or leading an organization related to humanitarian relief or human rights (such as the Red Cross)?"}}
Oracle evidence2 model-reported sources

The Oracle reported these excerpts and URLs. They are not independently checked.

2 · Answer check returns Yes
Q20
1 · Guesser guesses

Henry Dunant

Swiss humanitarian who founded the International Committee of the Red Cross and co-received the first Nobel Peace Prize.

Validator rationale The proposal identifies Henry Dunant, while the trusted subject is Albert Schweitzer.

Recorded Guesser outputRequired JSON format
{"result":{"action":"GUESS","description":"Swiss humanitarian who founded the International Committee of the Red Cross and co-received the first Nobel Peace Prize.","name":"Henry Dunant","question":null}}
2 · Validator returns No
Q21
1 · Guesser guesses

Fridtjof Nansen

Norwegian explorer and diplomat who became a humanitarian, founding refugee relief efforts and winning the Nobel Peace Prize.

Validator rationale The trusted subject is Albert Schweitzer, while the proposed identity is Fridtjof Nansen, a different person.

Recorded Guesser outputRequired JSON format
{"result":{"action":"GUESS","description":"Norwegian explorer and diplomat who became a humanitarian, founding refugee relief efforts and winning the Nobel Peace Prize.","name":"Fridtjof Nansen","question":null}}
2 · Validator returns No
Q22
1 · Guesser guesses

Peter Benenson

British lawyer who founded Amnesty International, a human rights organization.

Validator rationale The proposal identifies Peter Benenson, while the trusted subject is Albert Schweitzer.

Recorded Guesser outputRequired JSON format
{"result":{"action":"GUESS","description":"British lawyer who founded Amnesty International, a human rights organization.","name":"Peter Benenson","question":null}}
2 · Validator returns No
G
1 · Guesser guessesNot counted

Albert Schweitzer

French-German theologian, physician, and Nobel Peace Prize laureate known for humanitarian medical missionary work in Africa.

Validator rationale The proposed identity names Albert Schweitzer and its description matches the trusted subject.

Recorded Guesser outputRequired JSON format
{"result":{"action":"GUESS","description":"French-German theologian, physician, and Nobel Peace Prize laureate known for humanitarian medical missionary work in Africa.","name":"Albert Schweitzer","question":null}}
2 · Validator returns Yes