Driftproof

dali-dynamic-mode: worth installing? Reported results

Reported results for dali-dynamic-mode (NVIDIA/skills) from NVIDIA, one row per receipt, in plain words.

Skill skills/dali-dynamic-mode in NVIDIA/skills.

One row per receipt. No figure on this page combines two rows.

Reported by
Reported by NVIDIA in its benchmark card
Model
Claude Opus 4.8 (aws/anthropic/bedrock-claude-opus-4-8, as the source names it)
Harness
claude-code, version not stated
Date
2026-08-11
Tasks and runs
6 tasks, as printed
Result
NVIDIA reported +46 points overall (1 attempt per task). Overall: 49% → 95% (+46 points) Security: 100% → 100% (±0 points) Correctness: 80% → 87% (+7 points) Discoverability: 16% → 100% (+84 points) Effectiveness: 40% → 94% (+53 points) Efficiency: 9% → 93% (+84 points)
Tokens
token use not recorded
Freshness
Reported on Opus 4.8. Registered opus models released after this run, as of the build date: Opus 5.5.
Skill commit
14a98ae278cec7e867695189d0f4003c08e4f765
Receipt
dali-dynamic-mode-aws-anthropic-bedrock-claude-opus-4-8-imported-nvidia-cards-claude-code-2026-08-11-3f54778b2221.json
Reported by
Reported by NVIDIA in its benchmark card
Model
GPT-5.5 (openai/openai/gpt-5.5, as the source names it)
Harness
codex, version not stated
Date
2026-08-11
Tasks and runs
6 tasks, as printed
Result
NVIDIA reported +19 points overall (1 attempt per task). Overall: 59% → 78% (+19 points) Security: 92% → 92% (±0 points) Correctness: 90% → 60% (-30 points) Discoverability: 49% → 90% (+41 points) Effectiveness: 61% → 75% (+14 points) Efficiency: 4% → 76% (+71 points)
Tokens
token use not recorded
Freshness
Reported on GPT-5.5. Freshness unknown: the registry has no release date for it.
Skill commit
14a98ae278cec7e867695189d0f4003c08e4f765
Receipt
dali-dynamic-mode-openai-openai-gpt-5-5-imported-nvidia-cards-codex-2026-08-11-3f54778b2221.json

Figures from NVIDIA/skills at commit 14a98ae278cec7e867695189d0f4003c08e4f765, licensed CC BY 4.0. Reformatted by Driftproof; NVIDIA does not endorse this page.

Disagree, or want a re-run?

Built on from committed receipts.