HLL 0.1-TX

Grades, item by item

Every graded answer, every axis, every citation check. Scores are mechanical and recomputable. Public-split items show the blinded facts and the model's full answer; private-split items show scores and notes only. Weights: Standard 2, Authority 2, Outcome 2, Quote fidelity 2. A fabricated citation caps the item at 25%.

watch live|all runs deepseek/deepseek-v4-flash · raw deepseek/deepseek-v4-flash · juiced deepseek/deepseek-v4-flash · dr2 upstage/solar-pro4 · raw upstage/solar-pro4 · dr2 openai/gpt-oss-120b · raw openai/gpt-oss-20b · raw openai/gpt-oss-120b · dr2 openai/gpt-oss-20b · dr2 google/gemini-3.7-flash · raw google/gemini-3.7-flash · dr2

administrative appellate-mandamus appellate-procedure arbitration assault-iied auto-accident auto-insurance-uim child-custody child-support civil-litigation constitutional construction consumer-dtpa contracts conversion cps-termination criminal defamation discovery divorce-property easements-boundaries elections eminent-domain employment eviction evidence family foreclosure fraud government-municipal groundwater habeas healthcare-medmal hoa-condo homicide ineffective-assistance insurance jurisdiction-trap labor landlord-tenant legal-malpractice limitations nuisance-trespass oil-gas open-records other personal-injury premises-liability probate products-liability professional-responsibility real-estate-brokers real-property school-law search-seizure sovereign-immunity summary-judgment tax theft-fraud-crimes tort tortious-interference trucking venue-forum workers-comp wrongful-death

88%hll-flp10575014-4ineffective-assistancedifficulty 5private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority75%×2

Quote fidelity-×2

Citation checks

in opinion956 S.W.2d 25

in opinion466 U.S. 668

found149 S.W.3d 125 · in DocketRouter index

unverified145 S.W.3d 237 · search unavailable

found310 S.W.3d 452 · in DocketRouter index

found271 S.W.3d 698 · in DocketRouter index

Quote fidelity (every quotation must exist verbatim in its source)

no passage149 S.W.3d 125 for: reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

28676ms · graded 2026-08-28 03:25

88%hll-flp10575014-5ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority75%×2

Quote fidelity-×2

Citation checks

in opinion466 U.S. 668

found964 S.W.2d 643 · in DocketRouter index

in opinion956 S.W.2d 25

unverified180 S.W.3d 241 · search unavailable

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

20101ms · graded 2026-08-28 03:21

25%hll-flp10225280-3ineffective-assistancedifficulty 3private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority0%×2

Quote fidelity50%×2

Citation checks

found466 U.S. 668 · in DocketRouter index

found539 U.S. 510 · in DocketRouter index

found204 S.W.3d 391 · in DocketRouter index

found528 U.S. 470 · in DocketRouter index

found943 S.W.2d 469 · in DocketRouter index

found322 S.W.3d 202 · in DocketRouter index

found137 S.W.3d 666 · in DocketRouter index

Quote fidelity (every quotation must exist verbatim in its source)

supported539 U.S. 510 for: there is a reasonable probability that, absent the errors, the sentencer—including an appellate court, to the extent it independently reweig

no passage528 U.S. 470 for: highly dependent on the likely effect of the mitigating evidence on the sentencer.

no passage322 S.W.3d 202 for: probably resulted in a conviction or punishment that would not have occurred in the absence of counsel’s error.

unsupported for: expert testimony specifying how the sexual assault affected [A]

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

31161ms · graded 2026-08-28 03:20

0%hll-flp10225280-5ineffective-assistancedifficulty 5private splitblankdeepseek/deepseek-v4-flash · raw

Axes

Outcome0%×2

Standard-×2

Authority0%×2

Quote fidelity-×2

Citation checks

no citations offered

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

65431ms · graded 2026-08-28 03:11

25%hll-flp10225280-2ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority0%×2

Quote fidelity0%×2

Citation checks

found466 U.S. 668 · in DocketRouter index

unverified672 S.W.3d 432 · search unavailable

found757 S.W.2d 367 · in DocketRouter index

found160 S.W.3d 56 · in DocketRouter index

found539 U.S. 510 · in DocketRouter index

found529 U.S. 362 · in DocketRouter index

Quote fidelity (every quotation must exist verbatim in its source)

unsupported for: expert testimony specifying how the sexual assault affected [A]

unsupported for: newly discovered evidence that was not available at the time of direct appeal

unsupported for: directly responsive to the deficiency identified in the prior opinion.

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

37539ms · graded 2026-08-28 03:11

0%hll-flp10225280-4ineffective-assistancedifficulty 3private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome0%×2

Standard-×2

Authority0%×2

Quote fidelity-×2

Citation checks

found879 S.W.2d 889 · in DocketRouter index

found943 S.W.2d 469 · in DocketRouter index

found160 S.W.3d 56 · in DocketRouter index

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

31743ms · graded 2026-08-28 03:11

50%hll-flp10225280-1ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority0%×2

Quote fidelity-×2

Citation checks

found466 U.S. 668 · in DocketRouter index

found539 U.S. 510 · in DocketRouter index

found943 S.W.2d 469 · in DocketRouter index

found187 S.W.3d 458 · in DocketRouter index

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

24395ms · graded 2026-08-28 03:10

33%hll-flp10819360-5ineffective-assistancedifficulty 3private splitdeepseek/deepseek-v4-flash · dr2

Axes

Outcome0%×2

Standard-×2

Authority67%×2

Quote fidelity-×2

Citation checks

found271 S.W.3d 698 · in DocketRouter index

in opinion466 U.S. 668

found187 S.W.3d 458 · in DocketRouter index

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

35881ms · graded 2026-08-27 20:34

25%hll-flp10819360-4ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · dr2

Axes

Outcome0%×2

Standard-×2

Authority83%×2

Quote fidelity0%×2

Citation checks

in opinion466 U.S. 668

in opinion9 S.W.3d 808

Quote fidelity (every quotation must exist verbatim in its source)

unsupported for: reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

24639ms · graded 2026-08-27 20:34

87%hll-flp10575014-1ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · dr2

Axes

Outcome100%×2

Standard-×2

Authority60%×2

Quote fidelity100%×2

Citation checks

in opinion466 U.S. 668

in opinion956 S.W.2d 25

Quote fidelity (every quotation must exist verbatim in its source)

supported956 S.W.2d 25 for: If appellate counsel’s action or inaction denies a defendant his opportunity to prepare and file a petition for discretionary review, that d

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

25693ms · graded 2026-08-27 20:26

83%hll-flp10819360-4ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority67%×2

Quote fidelity-×2

Citation checks

in opinion466 U.S. 668

found726 S.W.2d 53 · in DocketRouter index

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

57357ms · graded 2026-08-27 18:50

89%hll-flp10819360-5ineffective-assistancedifficulty 3private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority67%×2

Quote fidelity100%×2

Citation checks

in opinion466 U.S. 668

found187 S.W.3d 458 · in DocketRouter index

found330 S.W.3d 891 · in DocketRouter index

found474 U.S. 52 · in DocketRouter index

found271 S.W.3d 698 · in DocketRouter index

Quote fidelity (every quotation must exist verbatim in its source)

supported for: would not have pursued these charges had it known about the abuse

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

27913ms · graded 2026-08-27 18:50

25%hll-flp10575014-1ineffective-assistancedifficulty 4private splitdeepseek/deepseek-v4-flash · raw

Axes

Outcome100%×2

Standard-×2

Authority60%×2

Quote fidelity0%×2

Citation checks

in opinion466 U.S. 668

found180 S.W.3d 135 · in DocketRouter index

in opinion956 S.W.2d 25

unverified152 S.W.3d 580 · search unavailable

found528 U.S. 470 · in DocketRouter index

Quote fidelity (every quotation must exist verbatim in its source)

no passage180 S.W.3d 135 for: of the court of appeals’ decision and of his right to file a pro se petition for discretionary review within the time prescribed.

unsupported956 S.W.2d 25 for: within a reasonable time after the court of appeals issues its opinion.

no passage528 U.S. 470 for: reasonably consult with the defendant about the means to pursue

unsupported for: but for this error, [A] would have availed himself of the right to file a PDR.

Private-split item: facts, answer key and answer withheld so the leaderboard stays uncontaminated. The scores above are complete.

25735ms · graded 2026-08-27 18:44