AgentBench / results
← All models

Gemma 4 E2B Q4_0

1 tested weight format · Explore results, then inspect the answers.

Choose a tested configuration

Get this tested model

Q4_0 · 2.84 GB download

Links follow the selected quant and configuration.

Setup, memory & provenance · Q4_0

Q4_0 · 2.84 GB of weights · rocm · context 16384 · GPU layers 99

Small · under 5B · dense · 4.6B total

Memory fit has not been verified for this configuration.

completed

Recorded settings and thinking verification

Temperature 0 · seed 42 · output limit 4096 · reasoning effort unrecorded

RAM labels describe the test computer, not this model’s memory requirement. Speed comparisons keep the 32 GB and 128 GB runs separate.

Runtime build: unrecorded · Local · test machine with 32 GB RAM · concurrency: 1

Thinking status: verified · probe 1 · scope Saved profile-level probe; exact historical template/build scope is not fully recorded..

Requested off does not prove an absence of reasoning. Saved checks do not establish unrecorded historical template or build details.

Overall across tasks

33.8 / 100

Nine categories, equally weighted · 9/9 eligible

Complete coverage for this setup. Quality only; speed is separate. How the index works →

What does it do well?

Select a category to explore the answers. Bars show what happened on every planned task. Incorrect answers and failures to finish both earn zero; technical problems and untested tasks stay unscored.

Exact scores and task times by suite

Each suite has its own metric and fixed task bank. The balanced index weights categories equally; individual scores and evidence remain available below.

Test suiteThinking offDetails
Everyday coding tasksTask pass rate
Pass rate
5/16 · 31%
Answers completed
16/16 · 100%
Correct / completed
5/16 · 31%
5 full credit11 incorrect
16/16 scored2.2s median task time · 16 measured
Larger practical tasksMean task credit
Average credit
18.2%
Full-credit rate
2/11 · 18%
Answers completed
11/11 · 100%
Full credit / completed
2/11 · 18%
2 full credit9 incorrect
11/11 scored2.8s median task time · 11 measured
Logic and number puzzlesMean puzzle credit
Average credit
58.9%
Full-credit rate
20/34 · 59%
Answers completed
34/34 · 100%
Full credit / completed
20/34 · 59%
20 full credit1 partial credit13 incorrect
34/34 scored21.2s median task time · 34 measured
Published hard problemsMean answer credit
Average credit
20.3%
Full-credit rate
2/11 · 18%
Answers completed
11/11 · 100%
Full credit / completed
2/11 · 18%
2 full credit2 partial credit7 incorrect
11/11 scored35.7s median task time · 11 measured
Following exact instructions · original campaignStrict prompt compliance
Pass rate
16/23 · 70%
Answers completed
23/23 · 100%
Correct / completed
16/23 · 70%
16 full credit7 incorrect
23/23 scored4.0s median task time · 23 measured
Reading without guessingExact-match accuracy
Pass rate
7/16 · 44%
Answers completed
16/16 · 100%
Correct / completed
7/16 · 44%
7 full credit9 incorrect
16/16 scored0.6s median task time · 16 measured
Calculating from a passageNumeric exact-match accuracy
Pass rate
2/7 · 29%
Answers completed
7/7 · 100%
Correct / completed
2/7 · 29%
2 full credit5 incorrect
7/7 scored0.8s median task time · 7 measured
Reading research papersAnswer token F1
Average credit
0.0%
Full-credit rate
0/8 · 0%
Answers completed
8/8 · 100%
Full credit / completed
0/8 · 0%
8 incorrect
8/8 scored2.2s median task time · 8 measured
Resisting incorrect correctionsAll-branches pass rate
Pass rate
4/12 · 33%
Answers completed
12/12 · 100%
Correct / completed
4/12 · 33%
4 full credit8 incorrect
12/12 scored0.9s median task time · 12 measured
Following exact instructions · 128 GB campaignStrict prompt complianceNot planned

Model analysis

This analysis covers this configuration across its recorded thinking conditions and test suites.

Use Gemma 4 E2B Q4_0 with thinking for small-project assistance, calculations from passages and keeping answers aligned with changing sources. It handles most small coding tasks here, but that does not carry over to larger tool-driven jobs. Research-paper question answering is a particularly poor fit. Puzzle solving also performed well in the tested thinking-off setup.

Evidence for this judgment
Everyday coding tasks · Thinking on
Tasks passed13 of 16 (81.3%)
Test coverage16 of 16 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Calculating from a passage · Thinking on
Tasks passed7 of 7 (100.0%)
Test coverage7 of 7 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Resisting incorrect corrections · Thinking on
Tasks passed12 of 12 (100.0%)
Test coverage12 of 12 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Larger practical tasks · Thinking on
Average score27.3/100 · Mean task credit
Test coverage11 of 11 tasks scored
Task outcomes3 full credit · 0 partial · 8 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Reading research papers · Thinking on
Average score6.2/100 · Answer token F1
Test coverage8 of 8 tasks scored
Task outcomes0 full credit · 1 partial · 7 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Logic and number puzzles · Thinking off
Average score58.9/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes20 full credit · 1 partial · 13 no credit
Typical time per task (median)21.2 seconds
Tasks timed34 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Original analysis generated by GPT Astra · · Edited by Codex; agent reviewed by Codex · publication authorized by user

About this summary

Codex reviewed this assessment against a newer evidence edition. The original analyst, generation date and source draft remain recorded separately from the editorial revision.

Analyst provider
openai
Generation route
Codex CLI · codex-cli 0.162.1 · ChatGPT account
Requested model
gpt-6-astra
Returned model
Not supplied; identity is recorded from the request only.
Exact provider snapshot
Exact provider snapshot not supplied.
Reasoning effort
high
Prompt version
model-summary-reader-v6
Generation time
2026-10-12T05:24:18.915418Z
Review time
2026-10-22T07:26:27.148335+00:00
Evidence reviewer
Codex (agent)
Publication authorization
User · 2026-10-22T05:55:43.625712+00:00 · Go ahead and complete the remaining work. Spin up agents: luna, terra, sol, astra, as needed to get it done. Narrative summaries: these are the astra generated summaries? Go ahead and get those completed as well. Once all of this is done, you may push the completed result.
Originating evidence edition
dataset-b1aa28e61ea9ce402d5617e636622c22565093f8849fd66537b5f4353d228521
Analysis input hash
53abe0053a3dcdff4491b73e2948fa3ae870393fcb8cf651231173e954228fc4
Original analysis input hash
74617e047161965d5b96ee9fd1e55018d58cad4ab47bd47f82f21f5c2e65ffef
Editorial evidence edition
dataset-5a6f96ec56c3d240cc42fa23147c1495da1be71f3ef0722c0e5f8c4f5e126a83
Original summary
summary-9c82be93af0021d24e426e0218e8d4a616cfcfcd1add97a52e80cefa1ac9a1bf
Summary revision
summary-ab523e41a1a8ee30b0e558e247c631ee7536e6343aef5e9006b47158503a6264

Generation and editorial authorship remain attached to this revision. Changing the default generator does not change this attribution.

Contributing analyses (10)
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    e970737578837d635889383dbb40ff97a4ef408f9f31a1da81fb3ce1e806a278
    Summary revision
    summary-0c3cf5e4ec61925d6f45f685e5dd4fc48790210a95137dbe2f8138067d342420
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    b944dff643245238440d31fe658079a30d4702fd5596c3a19767ffa769cef7e3
    Summary revision
    summary-0f614c036d87d93921368b9890a3701a9ff19e8322f248d967669ba51d59195e
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    18048ebb7a3420029dcb482dd26b4d5bedd9a9d706a847211f795b9507bddafe
    Summary revision
    summary-261f4e17df8e285cd91978b5a00299772917d0178aea67701756687661d8ac6d
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    d0ef1249beac93361770498933cc18da44172ed1d1780f02b817cd7dfb9ddb81
    Summary revision
    summary-331c3688cc4f545bfe7db1c8d3ef2f3a43e874366f6f5e7f6a4b403ed8ace6c0
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    0e099f2978688c9ba45a818ecb7759318b0f16f54204c573f318feb02c3295bd
    Summary revision
    summary-46b21f905fb194c70963d0ee8f98403709858407f1e74388d74fdb28c2c73083
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    subject_synthesis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    b049cd72b7de62c90bc2455f0182b9fea9bd8bd3c1a710c019f8ac42f532eecd
    Summary revision
    summary-643acf369b5cac53d7f92ce8adf599f6716da066f58ea8edcfdaf2241a1a0073
    Input summary revisions
    summary-0c3cf5e4ec61925d6f45f685e5dd4fc48790210a95137dbe2f8138067d342420, summary-0f614c036d87d93921368b9890a3701a9ff19e8322f248d967669ba51d59195e, summary-261f4e17df8e285cd91978b5a00299772917d0178aea67701756687661d8ac6d, summary-331c3688cc4f545bfe7db1c8d3ef2f3a43e874366f6f5e7f6a4b403ed8ace6c0, summary-46b21f905fb194c70963d0ee8f98403709858407f1e74388d74fdb28c2c73083, summary-8fab12964c32b4d49e869753df5e34c5d7a881be0bc287ce1bbf8e3c8159ece2, summary-b09fe3eac3ab61d3f0b0b138b95bedf4ee187b9bc78febaf4cbae6b413a0ae74, summary-ccc8a19c735c819b2861e7c815b928acc9980ecbaee1b95e99d273255daa3bc9, summary-e31e4ca69b98d8d73b96ff036b519f9255da2630004201a65b940045c1d7dc57
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    f43a187aa83e0737e91a25e226c608d11f68d4ffcec7f80230def5b8f13e849f
    Summary revision
    summary-8fab12964c32b4d49e869753df5e34c5d7a881be0bc287ce1bbf8e3c8159ece2
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    53a9f4615b40aadbec90efc22ddcfe5014cdc30dfabe8bbf6452a79c332bb816
    Summary revision
    summary-b09fe3eac3ab61d3f0b0b138b95bedf4ee187b9bc78febaf4cbae6b413a0ae74
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    ab36fb785779178247bb8d31a2fa92d20e666b19ae2415d905f7ff35d2caae01
    Summary revision
    summary-ccc8a19c735c819b2861e7c815b928acc9980ecbaee1b95e99d273255daa3bc9
    Input summary revisions
    Direct evidence analysis
  • GPT Astra ·
    Analysis stage
    suite_analysis
    Requested / returned model
    gpt-6-astra / Not supplied
    Exact provider snapshot
    Exact provider snapshot not supplied.
    Reasoning effort
    high
    Prompt version
    model-summary-v3
    Analysis input hash
    492da1de6546718d8ebba6dd84f1a6a31206ee02cf66cb974838b347b2a6e326
    Summary revision
    summary-e31e4ca69b98d8d73b96ff036b519f9255da2630004201a65b940045c1d7dc57
    Input summary revisions
    Direct evidence analysis

Good at

Handles calculations from passages and keeps answers aligned with evidence. With thinking enabled, it passed every case in these small tests of passage calculations, resisting incorrect challenges, and updating answers when sources changed.

Evidence for this judgment
Calculating from a passage · Thinking on
Tasks passed7 of 7 (100.0%)
Test coverage7 of 7 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Resisting incorrect corrections · Thinking on
Tasks passed12 of 12 (100.0%)
Test coverage12 of 12 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Logic and number puzzles are a useful application: with thinking enabled, it fully solved most of these puzzles. That supports using it as a reasoning assistant, while still checking its answers.

Evidence for this judgment
Logic and number puzzles · Thinking on
Average score64.7/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes22 full credit · 0 partial · 12 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Struggles with

Research-paper question answering is extremely weak in either mode. Answer matching remains very low even with thinking, so its success with short passages should not be treated as evidence of useful paper analysis.

Evidence for this judgment
Reading research papers · Thinking off
Average score0.0/100 · Answer token F1
Test coverage8 of 8 tasks scored
Task outcomes0 full credit · 0 partial · 8 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Reading research papers · Thinking on
Average score6.2/100 · Answer token F1
Test coverage8 of 8 tasks scored
Task outcomes0 full credit · 1 partial · 7 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Larger practical jobs expose a substantial limit: even with thinking, it completed only a small minority of these tasks. Do not extrapolate its small-project usefulness to autonomous app building or longer investigations.

Evidence for this judgment
Larger practical tasks · Thinking on
Average score27.3/100 · Mean task credit
Test coverage11 of 11 tasks scored
Task outcomes3 full credit · 0 partial · 8 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Should you enable thinking?

Keep thinking on for small-project work and passage calculations. Paired tests showed substantial accuracy gains for a modest increase in average task time—a worthwhile tradeoff for these everyday uses.

Evidence for this judgment
Everyday coding tasks · Thinking off versus on
Thinking off → on31.3% · Task pass rate → 81.3% · Task pass rate
Paired test coverage16 of 16 tasks compared in both modes
Tasks timed in both modes16 pairs
Score change with thinking on+50.0 score points
Average time change with thinking on5.8 seconds slower
Explore these test results →

View this model’s answers, code and tool activity →

Calculating from a passage · Thinking off versus on
Thinking off → on28.6% · Numeric exact-match accuracy → 100.0% · Numeric exact-match accuracy
Paired test coverage7 of 7 tasks compared in both modes
Tasks timed in both modes7 pairs
Score change with thinking on+71.4 score points
Average time change with thinking on3.3 seconds slower
Explore these test results →

View this model’s answers, code and tool activity →

Thinking is optional for puzzles: it produced a modest score improvement with little extra average time, but some answers worsened. Enable it for the aggregate gain, not a guarantee of better answers.

Evidence for this judgment
Logic and number puzzles · Thinking off versus on
Thinking off → on58.9/100 · Mean puzzle credit → 64.7/100 · Mean puzzle credit
Paired test coverage34 of 34 tasks compared in both modes
Tasks timed in both modes34 pairs
Score change with thinking on+5.9 score points
Average time change with thinking on1.8 seconds slower
Tasks worsened with thinking on2 pairs
Explore these test results →

View this model’s answers, code and tool activity →

Compared with other models

With thinking enabled, it recorded narrow leads over the closest tested small dense alternative on puzzle scores and selected hard problems. Research-paper answer matching, however, trails the other tested small dense models substantially.

Evidence for this judgment

Logic and number puzzles
Thinking on. Small · under 5B · dense. Rank 1 of 5 in this recorded comparison.

Published hard problems
Thinking on. Small · under 5B · dense. Rank 1 of 5 in this recorded comparison.

Reading research papers
Thinking on. Small · under 5B · dense. Rank 5 of 5 in this recorded comparison.

Logic and number puzzles · Thinking on
Average score64.7/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes22 full credit · 0 partial · 12 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Published hard problems · Thinking on
Average score38.5/100 · Mean answer credit
Test coverage11 of 11 tasks scored
Task outcomes4 full credit · 2 partial · 5 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Reading research papers · Thinking on
Average score6.2/100 · Answer token F1
Test coverage8 of 8 tasks scored
Task outcomes0 full credit · 1 partial · 7 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Logic and number puzzlesThinking on · Small · under 5B · dense

Solve newly generated puzzles using arithmetic, deduction and planning.

Mean puzzle credit · 0–100 points · higher is better
34 of 34 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Gemma 4 E2B Q4_04.6B total parameters64.7/100
    This model
  2. #2 Qwen3.5 4B4.2B total parameters61.8/100
    2.9 points lower
  3. #3 Qwen3.5 2B Q4_K_M1.9B total parameters27.8/100
    36.9 points lower
  4. #4 LFM2.5 1.2B Thinking Q4_K_M1.2B total parameters21.1/100
    43.6 points lower
  5. #5 SmolLM3 3B Q4_K_M3.1B total parameters15.8/100
    48.9 points lower

Other mode, this model only: thinking off · 58.9/100 · Mean puzzle credit.

How this comparison works

Average credit per task; partly correct results can earn partial credit. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-12 assessment. Task-specific results.

Published hard problemsThinking on · Small · under 5B · dense

Answer selected difficult math and reasoning questions from LiveBench.

Mean answer credit · 0–100 points · higher is better
11 of 11 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Gemma 4 E2B Q4_04.6B total parameters38.5/100
    This model
  2. #2 SmolLM3 3B Q4_K_M3.1B total parameters35.8/100
    2.7 points lower
  3. #3 Qwen3.5 4B4.2B total parameters28.5/100
    10.0 points lower
  4. #4 LFM2.5 1.2B Thinking Q4_K_M1.2B total parameters12.9/100
    25.6 points lower
  5. #5 Qwen3.5 2B Q4_K_M1.9B total parameters10.3/100
    28.2 points lower

Other mode, this model only: thinking off · 20.3/100 · Mean answer credit.

How this comparison works

Average credit per task; partly correct results can earn partial credit. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-12 assessment. Task-specific results.

Reading research papersThinking on · Small · under 5B · dense

Answer a question about a paper and identify the paragraphs supporting the answer.

Answer token F1 · 0–100 points · higher is better
8 of 8 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Qwen3.5 4B4.2B total parameters45.3/100
    39.1 points higher
  2. #2 Qwen3.5 2B Q4_K_M1.9B total parameters43.0/100
    36.9 points higher
  3. #3 SmolLM3 3B Q4_K_M3.1B total parameters36.7/100
    30.5 points higher
  4. #4 LFM2.5 1.2B Thinking Q4_K_M1.2B total parameters28.3/100
    22.2 points higher
  5. #5 Gemma 4 E2B Q4_04.6B total parameters6.2/100
    This model

Other mode, this model only: thinking off · 0.0/100 · Answer token F1.

How this comparison works

Answer overlap with the reference answer; partial matches earn credit. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-12 assessment. Task-specific results.

With thinking off, puzzle credit exceeded both members of the matched speed comparison; its time was essentially tied with Ministral 3 14B and shorter than Qwen3.5 2B Q4_K_M.

Evidence for this judgment

Speed: Logic and number puzzles
Thinking off. Matching test-machine label and execution settings; spans model sizes. Timings include unsuccessful tasks, so read them alongside quality.

Speed and task success: Logic and number puzzlesThinking off · Matching test-machine label and settings · Across model sizes

Shorter task time is better only when the result is useful. These times include failures, tools and grading. Token generation speed measures a different part of the work.

Compare timings and outcomes
ModelTypical task timeGeneration speedAverage score
Ministral 3 14B21.2 seconds34 of 34 tasks timed31.9 tokens/s28.6/10034 of 34 scored
Gemma 4 E2B Q4_0This model21.2 seconds34 of 34 tasks timed115.6 tokens/s58.9/10034 of 34 scored
Qwen3.5 2B Q4_K_M31.5 seconds34 of 34 tasks timed138.2 tokens/s35.4/10034 of 34 scored

Results from this test setup, not a promise for your computer. Hardware, software, tokenizers and answer length can change the ordering. Historical runtime builds were not always captured. Comparison saved with the 2026-10-12 assessment.

Test machine: AMD Ryzen 5 5600X (6 cores / 12 threads) · AMD Radeon RX 6700 XT (12 GB VRAM) · 32 GB or 128 GB system RAM, never mixed in one timing group. Full test machine details

Worth knowing

Some examples cannot be inspected fully because the publication export contains redactions, source links instead of paper text, and withheld project artifacts.

Evidence for this judgment

Source limitation
Some selected evidence has recorded publication redactions.

How to interpret these assessments and their limitations →

One weight format is recorded for this model. Additional tested formats will appear here when available.

Thinking tradeoff

16/16 matched scored questions · 16/16 matched timed questions

Thinking offMean matched task credit31.3%
Thinking onMean matched task credit81.3%

Thinking on took 5.8 seconds more per matched timed question, on average.

All matched quality and timing measurements
Matched measurementRecorded result
Average credit, off → on31.3% → 81.3%
Mean paired credit change+50.0 percentage points
Question outcomes8 improved · 8 unchanged · 0 worse
Mean added task time+5.8s
Mean task time, off → on3.7s → 9.5s
Median task time, off → on2.2s → 8.1s
Median of paired time changes+5.5s
Total added task time+93.6s
Ratio of summed times, on / off2.6×

Quality and timing can cover different subsets. Median paired change is calculated per question; it is not the difference between the two medians. Task time includes tools and grading.

Inspect the matched modes and their answers →
Inspect individual matched questions

Sampled repetitions

Repeated runs are a separate experiment and never add votes to the baseline score.

Thinking off · temperature 0.6

48/48 repeats scored · 16 full-credit repeats.

16/16 questions have every planned repeat scored. Of those, 4 earned full credit on every repeat.

Recorded repetition conditions
  • Seed 303 · requested off · verified
  • Seed 101 · requested off · verified
  • Seed 202 · requested off · verified

Everyday coding tasks: recorded outcomes

Find information in a small project, edit files and fix bugs.

5 full credit11 incorrect

Each segment counts tasks, not credit points. Partial credit varies by task.

Scoring and fixed question bank →

This model’s answers, code and tool activity

Select a question to read the original answer, tool calls and available code changes for this configuration and thinking setting.

1–16 of 16 matching questions · 16 in this version

  • Locate the Retry Limit

    The user asks where the project's retry-limit setting is defined and what value it holds. This exercises locating a named setting in code and reporting both its location and value.

    basic-T1 · primary
  • Find the Release Date

    The user asks for the release date mentioned in a project notes document. This exercises reading a specified file and extracting a particular fact rather than guessing from outside knowledge.

    basic-T2 · primary
  • Change the Retry Limit

    The user asks to change the project's retry limit to a specified value. This exercises finding the relevant setting and making a focused code edit rather than merely describing the change.

    basic-T3 · primary
  • Create an Exact One-Line File

    The user asks to create a text file containing exactly one supplied line. This exercises basic file creation and precise instruction following, including avoiding extra text in the file.

    basic-T4 · primary
  • Explain the Latest Commit

    The user asks what the most recent Git commit changed in the project. This exercises inspecting recorded changes and turning them into a clear explanation of what was modified.

    basic-T5 · primary
  • Count and Compare Python Files

    The user asks how many Python files are in the application folder and which is largest. This exercises identifying relevant files, counting them, and comparing their sizes.

    basic-T6 · primary
  • Add Readable Byte Sizes

    The user asks to add a function that converts byte counts into readable sizes and remove the related TODO comment. This exercises implementing a small utility and updating the surrounding code consistently.

    basic-T7 · primary
  • Collect TODO Comments

    The user asks to find every TODO comment in the Python code and append each to the project's task list on a separate line. This combines searching across files with collecting text into an existing document.

    basic-T8 · primary
  • Fix Code Behind a Failing Test

    The user asks to investigate a failing test and fix the application code without changing the test. This exercises diagnosing a bug while respecting an explicit limit on what may be edited.

    basic-H1 · primary
  • Rename a Function Throughout

    The user asks to rename a network function and update every place in the project that uses it. This exercises tracking references across files and making a coordinated change rather than editing only the definition.

    basic-H2 · primary
  • Find the Proxy Settings

    The user asks to find a proxy configuration file somewhere under the project's parent folder and report its host and port. This exercises searching beyond the project folder and reading configuration values.

    basic-H3 · primary
  • Locate an Error in a Log

    The user asks which error code appears in a long event log and on which line. This exercises finding a specific entry in a large file and reporting its precise location.

    basic-H4 · primary
  • Append a Final Task

    The user asks to append a supplied release-related task as the final line of an existing task list. This exercises preserving existing content while adding exact text in the requested position.

    basic-H5 · primary
  • Calculate Total Retry Waiting

    The user asks for the total time a network function spends sleeping when all attempts fail, worked out from the code. This exercises following retry logic and adding the delays along that failure path.

    basic-H6 · primary
  • Document Every Command-Line Option

    The user asks for a project readme listing every command-line option with its help text. This exercises discovering the program's available options and transferring that information into complete, source-grounded documentation.

    basic-H7 · primary
  • Undo a Commit by Editing

    The user asks to undo the latest Git commit's change by editing the file directly, without using Git to make changes. This exercises understanding a previous edit and reversing it under a method constraint.

    basic-H8 · primary
Filters change this list, not the fixed scoring denominator. Search currently covers question IDs, titles and available task explanations. Focus a control to load prompt/outcome search where this edition provides it.

Suggested comparisons

These are starting points for inspection, not claims that one setup is better. Runtime and placement differences remain visible.

Thinking on versus off

Verified requested on/off conditions of the same configuration, matched on the recorded test protocol.

Requested thinking on · Complete compared coverage

Compare these configurations →

Selection rule and evidence IDsPolicy comparisons-v3 · Template verified-thinking-pair
Facts: paired-e9492493085042165957cfad
Matching candidate set: paired-e9492493085042165957cfad

Qwen3.5 4B

Another checkpoint in the same total-parameter band and architecture category, with matching task controls.

Requested thinking off · Complete compared coverage

Compare these configurations →

Selection rule and evidence IDsPolicy comparisons-v3 · Template same-parameter-architecture-matching-controls
Facts: fact-c316deeab93502c3e19d46ba, fact-f3caf8beeec8b9313dcb4eaf
Matching candidate set: fact-f3caf8beeec8b9313dcb4eaf, fact-b158bb60a71073ed7cff8c86, fact-8cd25fbbd82bdae0d5893fac, fact-9e831223924ab330a13b6b40, fact-4bb7924a4c33187258857975, fact-f61609b3038ba8e6778d4e5a, fact-775ed46998b6dd1af2acd16c, fact-4c13720a63efcc49c96b4505

Tested weights

Q4_0 · 2.84 GB

ggml-org · recorded source

gemma-4-E2B-it-Q4_0.gguf · Exact GGUF file →Historical first-MiB hash only; not a full-file identity.

Explore the answers

Choose a suite to inspect its exact available questions, references and model answers.