DEPTH method synthesizes deeper terminal tasks for agents
DEPTH tasks exceed existing datasets in depth and work.
Stories per day · 30 days
2 RSS
Relatedevaluation1huggingface1lifelong-agents1long-horizon-agents1memory1payments1retrieval1terminal-tasks1token-efficiency1tool-use1
6 stories, newest first · one card per story
Stories tagged #agent-benchmarks
TokenWorth scores memory worth for lifelong agents under tight budgets
Across three LifelongAgentBench settings it records the highest success while emitting the fewest output tokens of the methods tested.
BizBench tests long-horizon payment agents on constraint-valid execution
Claude Opus 5 and GPT-6 Sol reach only 64.0% and 61.3% strict success on 50 payment tasks.
GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills
GraphSkillEvo evolves graph-structured LLM agent skills with mutation and crossover, beating SkillOpt by 4.01% on GPT-5.4-nano.
Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
DualViewEval compresses agent benchmarks by jointly modeling outcome and process signals, achieving 24x-40x compression with only 20 tasks on APEX-Agents and BFCL.
GPT-6 Astra pilots a surveillance drone and runs a business on its own
GPT-6 Astra outperforms Claude Fable 5.1 on Vending-Bench and becomes the first model to beat the human-AI baseline on all five Drone-Bench subtasks.
1