Hilde · Decision models · answer key v2

Grammar · topic 11

Word endings and what they do

79 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 model98.7%93.2% to 99.8%78 of 79
GPT-6 Astrahigh reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
GPT-6.1 Solhigh reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
GPT-6 Astralow reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
GPT-6 Astralow reasoning · 1 question per request · Chat model98.7%93.2% to 99.8%78 of 79
GPT-6.1 Sollow reasoning · 1 question per request · Chat model98.7%93.2% to 99.8%78 of 79
GPT-6.1 Sollow reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
Claude Fable 5.1high reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
Claude Opus 5.5high reasoning · 16 questions per request · Chat model98.7%93.2% to 99.8%78 of 79
Claude Opus 5.5low reasoning · 1 question per request · Chat model97.5%91.2% to 99.3%77 of 79
Gemini 3.8 Flashlow reasoning · 1 question per request · Chat model96.2%89.4% to 98.7%76 of 79
Gemini 4 Argonlow reasoning · 16 questions per request · Chat model94.9%87.7% to 98.0%75 of 79
GPT-6.1 Terrahigh reasoning · 16 questions per request · Chat model94.9%87.7% to 98.0%75 of 79
DeepSeek-V4.1-Flash, first releasereasoning on · 16 questions per request · Chat model94.9%87.7% to 98.0%75 of 79
Claude Opus 5.5low reasoning · 16 questions per request · Chat model94.9%87.7% to 98.0%75 of 79
Claude Sonnet 5.5high reasoning · 1 question per request · Chat model94.9%87.7% to 98.0%75 of 79
Claude Sonnet 5.5high reasoning · 16 questions per request · Chat model94.9%87.7% to 98.0%75 of 79
Grok 4.7high reasoning · 16 questions per request · Chat model93.7%86.0% to 97.3%74 of 79
GPT-6.1 Terralow reasoning · 16 questions per request · Chat model92.4%84.4% to 96.5%73 of 79
Gemini 3.8 Flashlow reasoning · 16 questions per request · Chat model91.1%82.8% to 95.6%72 of 79
GLM-5.3-Flashreasoning on · 16 questions per request · Chat model88.6%79.7% to 93.9%70 of 79
Kimi K3.1 Prolow reasoning · 16 questions per request · Chat model81.0%71.0% to 88.1%64 of 79
Qwen3.8-27Blow reasoning · 16 questions per request · Chat model79.7%69.6% to 87.1%63 of 79
GLM-5.3low reasoning · 16 questions per request · Chat model78.5%68.2% to 86.1%62 of 79
Liquid d1-ultra1 question per request · Decision model68.4%57.5% to 77.6%54 of 79
Gemma 4 31Breasoning off · 16 questions per request · Chat model68.4%57.5% to 77.6%54 of 79
Fastino GLiDE v1.2local · 1 question per request · Decision model65.8%54.8% to 75.3%52 of 79
Kimi K3.1 Proreasoning none · 16 questions per request · Chat model64.6%53.6% to 74.2%51 of 79
GPT-6 Lunalow reasoning · 16 questions per request · Chat model64.6%53.6% to 74.2%51 of 79
Drex Max v11 question per request · Decision model64.6%53.6% to 74.2%51 of 79
Drex Max 1.11 question per request · Decision model64.6%53.6% to 74.2%51 of 79
Mammoth-11 question per request · Decision model64.6%53.6% to 74.2%51 of 79
Perplexity Decider 27B v1.11 question per request · Decision model63.3%52.3% to 73.1%50 of 79
Fastino GLiDE1 question per request · Decision model62.0%51.0% to 71.9%49 of 79
vLLM-SR d31 question per request · Decision model62.0%51.0% to 71.9%49 of 79
vLLM-SR d3-ultratransformers, rev 52ba9e7 · 1 question per request · Decision model62.0%51.0% to 71.9%49 of 79
Perplexity Decider 27B v1.2local · 1 question per request · Decision model62.0%51.0% to 71.9%49 of 79
Gemma 4 31Breasoning off · 1 question per request · Chat model60.8%49.7% to 70.8%48 of 79
SubQ Snap Preview1 question per request · Decision model60.8%49.7% to 70.8%48 of 79
SubQ Snap 1.11 question per request · Decision model59.5%48.5% to 69.6%47 of 79
OneJev 27B1 question per request · Decision model59.5%48.5% to 69.6%47 of 79
Open-Jev 27B v1.11 question per request · Decision model59.5%48.5% to 69.6%47 of 79
DeepSeek-V4.1-Flash, first releasereasoning off · 16 questions per request · Chat model58.2%47.2% to 68.5%46 of 79
Perplexity Decider 27B v11 question per request · Decision model57.0%46.0% to 67.3%45 of 79
Claude Sonnet 5.5low reasoning · 16 questions per request · Chat model53.2%42.3% to 63.8%42 of 79
Liquid d1.11 question per request · Decision model51.9%41.1% to 62.6%41 of 79
Celeris-1 Decision1 question per request · Decision model51.9%41.1% to 62.6%41 of 79
Inception Mercury Decidevia OpenRouter · 1 question per request · Decision model51.9%41.1% to 62.6%41 of 79
Clef1 question per request · Decision model51.9%41.1% to 62.6%41 of 79
Jev 1.13.01 question per request · Decision model49.4%38.6% to 60.2%39 of 79
Liquid d11 question per request · Decision model49.4%38.6% to 60.2%39 of 79
Cygnet1 question per request · Decision model49.4%38.6% to 60.2%39 of 79
Qwen3.8-27Bstated confidence · reasoning off · 1 question per request · Chat model49.4%38.6% to 60.2%39 of 79
GPT-6 LunaDecisions API, beta · 1 question per request · Decision model48.1%37.4% to 58.9%38 of 79
Jev 1.14.01 question per request · Decision model48.1%37.4% to 58.9%38 of 79
GLM-5.3-Flashreasoning off · 16 questions per request · Chat model48.1%37.4% to 58.9%38 of 79
Winnow-12B1 question per request · Decision model48.1%37.4% to 58.9%38 of 79
Elephant-11 question per request · Decision model43.0%32.7% to 54.0%34 of 79
Drex 2.01 question per request · Decision model41.8%31.5% to 52.8%33 of 79
Qwen3.8-27Bvote share · reasoning off · 1 question per request · Chat model41.8%31.5% to 52.8%33 of 79
Perplexity Decider Flash v1.2local · 1 question per request · Decision model41.8%31.5% to 52.8%33 of 79
Clef-flash1 question per request · Decision model38.0%28.1% to 49.0%30 of 79
Erk-14Breasoning off · 16 questions per request · Chat model38.0%28.1% to 49.0%30 of 79
Qwen3.8-27Byes/no scoring · 1 question per request · Chat model36.7%26.9% to 47.7%29 of 79
Microsoft-Decision-1via OpenRouter · 1 question per request · Decision model36.7%26.9% to 47.7%29 of 79
Metask-Jev 4B1 question per request · Decision model32.9%23.6% to 43.9%26 of 79
Qwen3-14Breasoning off · 16 questions per request · Chat model30.4%21.3% to 41.2%24 of 79
meraGPT Decider 11 question per request · Decision model29.1%20.3% to 39.9%23 of 79
Liquid d1-2B (open weights)1 question per request · Decision model25.3%17.0% to 35.9%20 of 79
Strands Decider 2B1 question per request · Decision model24.1%16.0% to 34.5%19 of 79
GLiNER2.5 Multi (base)1 question per request · Chat model21.5%13.9% to 31.8%17 of 79
Laya multilingual1 question per request · Decision model20.3%12.9% to 30.4%16 of 79
GLiNER2.5-multi-Decide1 question per request · Decision model17.7%10.9% to 27.6%14 of 79
Laya1 question per request · Decision model16.5%9.9% to 26.1%13 of 79

Its questions

IDPageNumberAnswer keySetups rightDecision models right
35552271E56 of 7426 of 37
35562274C24 of 741 of 37
35572272A30 of 743 of 37
35582275E65 of 7431 of 37
35592273E49 of 7417 of 37
35602276D64 of 7429 of 37
35612287C62 of 7430 of 37
356222810C42 of 7415 of 37
356322811C46 of 7416 of 37
35642288C51 of 7420 of 37
356522812B55 of 7423 of 37
35662289D39 of 7416 of 37
356722813A54 of 7422 of 37
35682291B53 of 7421 of 37
35692295B63 of 7429 of 37
35702296B59 of 7428 of 37
35712292D23 of 742 of 37
35722293C64 of 7429 of 37
35732297E40 of 7416 of 37
35742294B58 of 7425 of 37
35752298C53 of 7422 of 37
35762309C47 of 7417 of 37
357723012A31 of 746 of 37
357823010B31 of 745 of 37
357923013E46 of 7418 of 37
358023014C51 of 7423 of 37
358123011A45 of 7416 of 37
35822311B29 of 746 of 37
35832315C67 of 7431 of 37
35842312D45 of 7416 of 37
35852316E36 of 7410 of 37
35862313A34 of 745 of 37
35872317D62 of 7430 of 37
35882314C53 of 7421 of 37
35892328E60 of 7429 of 37
359023211A63 of 7432 of 37
359123212B43 of 7414 of 37
35922329E42 of 7415 of 37
359323210A27 of 744 of 37
359423213C25 of 743 of 37
35952331A47 of 7418 of 37
35962334D68 of 7433 of 37
35972332D41 of 7415 of 37
35982335D40 of 7411 of 37
35992333B54 of 7423 of 37
36002336E57 of 7424 of 37
360123410C53 of 7422 of 37
36022347E68 of 7433 of 37
36032348A40 of 7414 of 37
360423411D52 of 7420 of 37
360523412E67 of 7433 of 37
36062349A52 of 7423 of 37
36072351A30 of 745 of 37
36082355E68 of 7433 of 37
36092352A36 of 7410 of 37
36102356A69 of 7432 of 37
36112353D26 of 741 of 37
36122357A40 of 7417 of 37
36132354E59 of 7428 of 37
36142368D18 of 740 of 37
361523612C62 of 7426 of 37
36162369B42 of 7412 of 37
361723613B29 of 742 of 37
361823610E20 of 743 of 37
361923614B40 of 7412 of 37
362023611B49 of 7422 of 37
36212371E46 of 7415 of 37
36222374C26 of 7416 of 37
36232372C54 of 7427 of 37
36242375E44 of 7412 of 37
36252373D66 of 7430 of 37
36262376E45 of 7415 of 37
36272387C41 of 7413 of 37
362823810A29 of 745 of 37
362923811B29 of 747 of 37
36302388E46 of 7417 of 37
363123812E59 of 7423 of 37
36322389E62 of 7427 of 37
363323813C48 of 7423 of 37