IHEval D
AI benchmark · Instruction Following
- Domains
- Instruction Following
- Tier
- D why
- Snapshot
What is IHEval?
IHEval measures whether language models follow the instruction hierarchy (system messages, user messages, conversation history, tool outputs), including when instructions at different priorities conflict.
General capability mostly
“a novel benchmark comprising 3,538 examples across nine tasks, covering cases where instructions in different priorities either align or conflict.”
arxiv.org/abs/2502.08745 ↗
Paper: arxiv.org/abs/2502.08745 ↗
Tier D
Long tail: little or no usage data yet.
How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.
- no usage data
- in curated RL directories (1 pt)
On X · unclear, low buzz, from 1 post about IHEval in the last month
IHEval shows up mainly as one reported metric in NVIDIA's Nemotron post-training comparisons, where it shows large gains, with little discussion of the benchmark itself.
How we got to tier D
- Adoption: no usage data. Tier D.
- On X: unclear, low buzz. “IHEval shows up mainly as one reported metric in NVIDIA's Nemotron post-training comparisons, where it shows large gains, with little discussion of the benchmark itself.” No move.
- In curated RL directories: 1 point, a floor of none.
- Confirmed: a line from its own site or paper places it.
A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.
Where IHEval sits
Its domain on the map, live. Hover a logo for its name; click to open it. Open it on the full map
Next: browse all benchmarks. Entry from the RL Research daily scrape of public sources, 2026-10-29 snapshot.