arXiv introduces LongBanana benchmark for long-context image synthesis
LongBanana contains 90 tasks with 4,670 expert-specified evaluation items for agent-driven multi-reference image generation.
Stories per day · 30 days
Relatedecosystem-drift1image-synthesis1long-context1multi-reference1skill-learning1steward1tool-use1visual-evidence1
2 stories, newest first · one card per story
Stories tagged #agent-benchmark
Δτ-Bench tests agents as tool ecosystems change
The strongest evaluated model or harness reached only 42.3% success on 1,588 tasks across six ecosystem states.