Hilde · Decision models · answer key v2

Grammar · topic 9

Verbs and their forms

91 scored questions. Accuracy by topic compares this topic with the others.

Every setup's result

Accuracy on this topic's questions, with its 95% interval. Few questions make wide intervals.

SetupAccuracyRight
Gemini 3.8 Flashhigh reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6 Astrahigh reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6.1 Solhigh reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6 Astralow reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6 Astralow reasoning · 1 question per request · Chat model100.0%95.9% to 100.0%91 of 91
Gemini 3.8 Flashlow reasoning · 1 question per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6.1 Sollow reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
Gemini 4 Argonlow reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
Claude Fable 5.1high reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
Gemini 3.8 Flashlow reasoning · 16 questions per request · Chat model100.0%95.9% to 100.0%91 of 91
GPT-6.1 Sollow reasoning · 1 question per request · Chat model97.8%92.3% to 99.4%89 of 91
Claude Opus 5.5high reasoning · 16 questions per request · Chat model97.8%92.3% to 99.4%89 of 91
GPT-6.1 Terrahigh reasoning · 16 questions per request · Chat model97.8%92.3% to 99.4%89 of 91
Claude Opus 5.5low reasoning · 1 question per request · Chat model97.8%92.3% to 99.4%89 of 91
DeepSeek-V4.1-Flash, first releasereasoning on · 16 questions per request · Chat model97.8%92.3% to 99.4%89 of 91
Grok 4.7high reasoning · 16 questions per request · Chat model95.6%89.2% to 98.3%87 of 91
GPT-6.1 Terralow reasoning · 16 questions per request · Chat model95.6%89.2% to 98.3%87 of 91
Claude Opus 5.5low reasoning · 16 questions per request · Chat model94.5%87.8% to 97.6%86 of 91
Claude Sonnet 5.5high reasoning · 1 question per request · Chat model94.5%87.8% to 97.6%86 of 91
Claude Sonnet 5.5high reasoning · 16 questions per request · Chat model86.8%78.4% to 92.3%79 of 91
Kimi K3.1 Prolow reasoning · 16 questions per request · Chat model86.8%78.4% to 92.3%79 of 91
GLM-5.3-Flashreasoning on · 16 questions per request · Chat model86.8%78.4% to 92.3%79 of 91
Liquid d1-ultra1 question per request · Decision model82.4%73.3% to 88.9%75 of 91
Qwen3.8-27Blow reasoning · 16 questions per request · Chat model80.2%70.9% to 87.1%73 of 91
GLM-5.3low reasoning · 16 questions per request · Chat model75.8%66.1% to 83.5%69 of 91
Gemma 4 31Breasoning off · 16 questions per request · Chat model75.8%66.1% to 83.5%69 of 91
Kimi K3.1 Proreasoning none · 16 questions per request · Chat model75.8%66.1% to 83.5%69 of 91
GPT-6 Lunalow reasoning · 16 questions per request · Chat model70.3%60.3% to 78.7%64 of 91
Claude Sonnet 5.5low reasoning · 16 questions per request · Chat model67.0%56.9% to 75.8%61 of 91
Gemma 4 31Breasoning off · 1 question per request · Chat model65.9%55.7% to 74.8%60 of 91
DeepSeek-V4.1-Flash, first releasereasoning off · 16 questions per request · Chat model65.9%55.7% to 74.8%60 of 91
GPT-6 LunaDecisions API, beta · 1 question per request · Decision model65.9%55.7% to 74.8%60 of 91
Fastino GLiDE1 question per request · Decision model64.8%54.6% to 73.9%59 of 91
Perplexity Decider 27B v1.11 question per request · Decision model64.8%54.6% to 73.9%59 of 91
Drex Max v11 question per request · Decision model63.7%53.5% to 72.9%58 of 91
vLLM-SR d31 question per request · Decision model62.6%52.4% to 71.9%57 of 91
Drex Max 1.11 question per request · Decision model62.6%52.4% to 71.9%57 of 91
Fastino GLiDE v1.2local · 1 question per request · Decision model61.5%51.3% to 70.9%56 of 91
Celeris-1 Decision1 question per request · Decision model61.5%51.3% to 70.9%56 of 91
vLLM-SR d3-ultratransformers, rev 52ba9e7 · 1 question per request · Decision model60.4%50.2% to 69.9%55 of 91
Perplexity Decider 27B v1.2local · 1 question per request · Decision model60.4%50.2% to 69.9%55 of 91
Perplexity Decider 27B v11 question per request · Decision model60.4%50.2% to 69.9%55 of 91
Cygnet1 question per request · Decision model59.3%49.1% to 68.9%54 of 91
Mammoth-11 question per request · Decision model58.2%48.0% to 67.8%53 of 91
Liquid d1.11 question per request · Decision model58.2%48.0% to 67.8%53 of 91
OneJev 27B1 question per request · Decision model58.2%48.0% to 67.8%53 of 91
Jev 1.14.01 question per request · Decision model58.2%48.0% to 67.8%53 of 91
SubQ Snap Preview1 question per request · Decision model57.1%46.9% to 66.8%52 of 91
Drex 2.01 question per request · Decision model57.1%46.9% to 66.8%52 of 91
Open-Jev 27B v1.11 question per request · Decision model57.1%46.9% to 66.8%52 of 91
Jev 1.13.01 question per request · Decision model56.0%45.8% to 65.8%51 of 91
SubQ Snap 1.11 question per request · Decision model54.9%44.7% to 64.8%50 of 91
GLM-5.3-Flashreasoning off · 16 questions per request · Chat model53.8%43.7% to 63.7%49 of 91
Clef1 question per request · Decision model51.6%41.5% to 61.6%47 of 91
Inception Mercury Decidevia OpenRouter · 1 question per request · Decision model50.5%40.5% to 60.6%46 of 91
Liquid d11 question per request · Decision model50.5%40.5% to 60.6%46 of 91
Winnow-12B1 question per request · Decision model50.5%40.5% to 60.6%46 of 91
Qwen3.8-27Bvote share · reasoning off · 1 question per request · Chat model46.2%36.3% to 56.3%42 of 91
Qwen3.8-27Bstated confidence · reasoning off · 1 question per request · Chat model45.1%35.2% to 55.3%41 of 91
Qwen3.8-27Byes/no scoring · 1 question per request · Chat model45.1%35.2% to 55.3%41 of 91
Microsoft-Decision-1via OpenRouter · 1 question per request · Decision model44.0%34.2% to 54.2%40 of 91
Perplexity Decider Flash v1.2local · 1 question per request · Decision model40.7%31.1% to 50.9%37 of 91
Metask-Jev 4B1 question per request · Decision model39.6%30.1% to 49.8%36 of 91
Elephant-11 question per request · Decision model36.3%27.1% to 46.5%33 of 91
Clef-flash1 question per request · Decision model36.3%27.1% to 46.5%33 of 91
Strands Decider 2B1 question per request · Decision model31.9%23.2% to 42.0%29 of 91
Qwen3-14Breasoning off · 16 questions per request · Chat model29.7%21.3% to 39.7%27 of 91
Erk-14Breasoning off · 16 questions per request · Chat model29.7%21.3% to 39.7%27 of 91
Liquid d1-2B (open weights)1 question per request · Decision model27.5%19.4% to 37.4%25 of 91
meraGPT Decider 11 question per request · Decision model25.3%17.5% to 35.1%23 of 91
GLiNER2.5-multi-Decide1 question per request · Decision model25.3%17.5% to 35.1%23 of 91
GLiNER2.5 Multi (base)1 question per request · Chat model17.6%11.1% to 26.7%16 of 91
Laya1 question per request · Decision model16.5%10.3% to 25.4%15 of 91
Laya multilingual1 question per request · Decision model15.4%9.4% to 24.2%14 of 91

Its questions

IDPageNumberAnswer keySetups rightDecision models right
8792027D67 of 7431 of 37
88020211D38 of 7416 of 37
8812028E50 of 7417 of 37
88220212D29 of 744 of 37
8832029D62 of 7427 of 37
88420213E59 of 7427 of 37
88520210A62 of 7428 of 37
8992051A63 of 7432 of 37
9002054E65 of 7432 of 37
9012055C70 of 7433 of 37
9022052D35 of 7411 of 37
9032056E52 of 7419 of 37
9042053D43 of 7415 of 37
9052057B58 of 7423 of 37
9142071B34 of 746 of 37
9152074A64 of 7431 of 37
9162075E48 of 7418 of 37
9172072E61 of 7430 of 37
9182073D33 of 747 of 37
9192076D37 of 748 of 37
9262091B33 of 7410 of 37
9272095D36 of 744 of 37
9282092A27 of 746 of 37
9292096A38 of 748 of 37
9302093C30 of 744 of 37
9312097B34 of 7414 of 37
9322094E36 of 748 of 37
9332108C47 of 7413 of 37
93421011B32 of 743 of 37
9352109E40 of 7411 of 37
93621012A53 of 7420 of 37
93721013C29 of 749 of 37
93821010D65 of 7431 of 37
9392111B16 of 741 of 37
9402114D62 of 7429 of 37
9412115B36 of 7412 of 37
9422112C48 of 7418 of 37
9432116A30 of 7412 of 37
9442113B30 of 742 of 37
9522131B58 of 7427 of 37
9532134E59 of 7427 of 37
9542135A63 of 7430 of 37
9552132D63 of 7429 of 37
9562136D64 of 7431 of 37
9572133C59 of 7425 of 37
9582147C68 of 7432 of 37
95921410E39 of 7411 of 37
96021411B56 of 7426 of 37
9612148D34 of 749 of 37
96221412A28 of 742 of 37
9632149E44 of 7418 of 37
34492011D66 of 7431 of 37
34502014B55 of 7423 of 37
34512015E71 of 7435 of 37
34522012E66 of 7432 of 37
34532016D52 of 7423 of 37
34542013C62 of 7429 of 37
34552031C49 of 7418 of 37
34562034C58 of 7422 of 37
34572035E71 of 7435 of 37
34582032D54 of 7423 of 37
34592036B44 of 7418 of 37
34602033B33 of 744 of 37
34612037D40 of 7414 of 37
34622048B68 of 7433 of 37
346320411E68 of 7432 of 37
34642049E65 of 7431 of 37
346520412C24 of 744 of 37
346620410D51 of 7420 of 37
346720413A21 of 744 of 37
34682068E62 of 7429 of 37
346920612B37 of 749 of 37
347020613E64 of 7430 of 37
34712069D69 of 7433 of 37
347220614A62 of 7430 of 37
347320610A25 of 742 of 37
347420615C57 of 7424 of 37
347520611C52 of 7421 of 37
347620810C47 of 7416 of 37
34772087C32 of 745 of 37
347820811E39 of 747 of 37
34792088B34 of 749 of 37
348020812A54 of 7423 of 37
34812089D44 of 7414 of 37
34822127E63 of 7432 of 37
348321211E45 of 7418 of 37
34842128D54 of 7420 of 37
348521212C38 of 7413 of 37
34862129A31 of 747 of 37
348721213D50 of 7420 of 37
348821210B60 of 7428 of 37