Skip to main content
PolicyBench

Qwen 3.8 Max

Alibaba · AI alone, no tools
Current board, snapshot 2026-11-11
United States rank
#47 of 52
US exact
74.7%
United States

United States benchmark

Exact match
74.7%
Bounded score
82.4%
Parse rate
100.0%
Eligibility flags
900/975

Exact match by programall 18 output groups, best first

Bar length is the unweighted exact-match rate on that program's outputs; the right column is the program's share of the headline weight, which follows household dollars, so the bars do not average to the headline.

  1. Local income tax100.0%0.3%
  2. Person-level Early Head Start eligibility100.0%3.1%
  3. Reduced-price school meals eligibility100.0%0.1%
  4. TANF98.0%0.3%
  5. Person-level Medicare eligibility97.7%11%
  6. SSI96.9%2.0%
  7. Free school meals eligibility95.0%0.7%
  8. Person-level CHIP eligibility92.1%0.2%
  9. Self-employment tax92.0%2.1%
  10. Person-level WIC eligibility89.3%0.3%
  11. Person-level Head Start eligibility86.8%1.2%
  12. Person-level Medicaid eligibility83.8%30%
  13. SNAP81.7%4.1%
  14. State refundable credits81.6%0.6%
  15. Payroll tax74.0%16%
  16. Federal refundable credits66.0%3.7%
  17. State tax before refundable credits57.0%6.0%
  18. Federal tax before refundable credits48.8%19%
Table view (exact, within 1%, outputs), hardest first
ProgramExactWithin 1%n
Federal tax before refundable credits48.8%48.8%84
State tax before refundable credits57.0%59.1%93
Federal refundable credits66.0%66.0%100
Payroll tax74.0%77.1%96
State refundable credits81.6%81.6%98
SNAP81.7%81.7%93
Person-level Medicaid eligibility83.8%83.8%173
Person-level Head Start eligibility86.8%86.8%38
Person-level WIC eligibility89.3%89.3%177
Self-employment tax92.0%92.0%100
Person-level CHIP eligibility92.1%92.1%177
Free school meals eligibility95.0%95.0%100
SSI96.9%96.9%97
Person-level Medicare eligibility97.7%97.7%172
TANF98.0%98.0%100
Local income tax100.0%100.0%98
Person-level Early Head Start eligibility100.0%100.0%38
Reduced-price school meals eligibility100.0%100.0%100

Hardest casesworst misses on positive references

Scores are from the frozen manuscript snapshot under the AI-alone condition: structured responses over the same household facts (whole-scenario or in output subsets, per the serving-configuration table), no tools, graded against PolicyEngine reference outputs. See the leaderboard and paper for methodology. The frozen US annotations cover 9,132 rows whose legacy threshold score is below 1. That universe contains 9,128 of 9,128 exact-match misses and 4 exact hits. Another 2,497 rows with bounded score below 100 were not selected and have no audit annotation.