PRISM Instruction Order Sensitivity
Models solve the same math problems under different orderings of fixed instructions. The evaluation compares accuracy across orderings on GSM8K and MATH-500.
- Made by
- Blessings Mambwe
- Released
- August 2026
- Kind
- Following precise instructions
- Level
- School
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Subject
- Mathematics
- Language
- English
- Interaction
- Single turn
- Items from
- Another benchmark
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
-
Qwen3-30B-A3B-Instruct Jul 2025 #144 on the capability index
- Last model added
- September 2025
- Scores refreshed
- November 20th, 2026
Qwen3-30B-A3B-Instruct leads with 66% Qwen3-30B-A3B-Instruct leads with 66%
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
- 1
Qwen3-30B-A3B-Instruct Jul 2025, GSM8K, 100 random orderings 66%
- 2
Gemini 2.5 Flash-Lite Sep 2025, GSM8K, 100 random orderings 16%
- 3 meta-llama/llama-3.2-3b-instruct GSM8K, 100 random orderings 0%
Source: Official PRISM Instruction Order Sensitivity leaderboard
Source: Official PRISM Instruction Order Sensitivity leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 66% | GSM8K, 100 random orderings | Jul 2025 | Official leaderboard |
| 2 | | 16% | GSM8K, 100 random orderings | Sep 2025 | Official leaderboard |
| 3 | meta-llama/llama-3.2-3b-instruct | 0% | GSM8K, 100 random orderings | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 66% | GSM8K, 100 random orderings | Jul 2025 | Official leaderboard |
| 2 | | 16% | GSM8K, 100 random orderings | Sep 2025 | Official leaderboard |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official PRISM Instruction Order Sensitivity leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Instructions and chat
- IFEval In use Models answer about 500 prompts containing verifiable instructions such as word limits or required keywords. Code checks each instruction for strict and loose accuracy.
Gemma 4 31B IT 98.9%
- IFBench In use 58 new out-of-domain verifiable output constraints. Scored by programmatic constraint checks (strict or loose). Flint 89.2%
- MultiChallenge In use Multi-turn conversations testing instruction retention, inference memory, versioned editing and self-coherence. Graded by LLM judge with rubrics. GPT-6.1 Sol 82%
- Multi-IF In use Models follow verifiable instructions over three-turn conversations in eight languages, extending IFEval. Scored by instruction accuracy at each turn. LFM3-1.7B 89.1%
- AdvancedIF In use Models respond to over 1,600 prompts with complex, multi-turn, and system-level instructions. A judge model checks each response against expert-written rubrics.
MAI-Thinking-1 85%
- MM-IFEval In use Multimodal models follow image-grounded instructions with exact output constraints. Checks mix rule-based verification and a judge model. GPT-4o 64.6%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 14 months ago, in September 2025, for Gemini 2.5 Flash-Lite. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (3 entries).
- Note Models scored Only 2 models we track, plus 1 other entry on its leaderboard.
- Warning Newest score The newest came 14 months ago, in September 2025, for Gemini 2.5 Flash-Lite.
- Note Frontier models None of the top 10 has a score. The strongest tested is Qwen3-30B-A3B-Instruct (#144).
- Pass Headroom The best is 66%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on github.com: 2 rows for models we track and 1 other entry at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.