Prime Intellect releases General-Agent, a synthetic reinforcement learning environment with self-evolving tasks
Prime Intellect introduced General-Agent, a fully synthetic reinforcement learning environment whose tasks self-evolve and grow in difficulty. The release includes 4,504 tool-use tasks spanning 1,040 domains and 8,159 unique tools. A synthesizer designs new tasks using databases and tools, while a solver runs multiple rollouts to measure pass rates. Tasks are accepted or escalated based on target performance bands. The system supports parallel execution of multi-hour agent episodes across broad task compositions.
#707Vincent Weisser@VINCENTWEISSERAutomating RL environments is the next step toward automating everything else.
Introducing general-agent by @mikasenghaas > open agentic environments with 1000s of tools are scarce, so we're building one that builds itself > A synthesizer evolves tasks across difficulty tiers, empirically gated by a solver. Hard tiers seed the next wave, hillclimbing toward frontier-level difficulty. > 4,504 tasks / 1,040 domains / 8,159 unique tools
Prime Intellect@PrimeIntellectThe next step toward automating AI is automating RL environments Introducing General-Agent: A fully synthetic environment whose task corpus self-evolves and grows harder over time 4,504 tool-use tasks · 1,040 domains · 8,159 unique tools
@mikasenghaas https://www.primeintellect.ai/blog/general-agent
Vincent Weisser@vincentweisserAutomating RL environments is the next step toward automating everything else. Introducing general-agent by @mikasenghaas > open agentic environments with 1000s of tools are scarce, so we're building one that builds itself > A synthesizer evolves tasks across difficulty tiers, empirically gated by a solver. Hard tiers seed the next wave, hillclimbing toward frontier-level difficulty. > 4,504 tasks / 1,040 domains / 8,159 unique tools
#1153Florian Brand@XEOPHONmaking the tech that closed labs have open auf giving it to everyone, one release after another :)
Prime Intellect@PrimeIntellectThe next step toward automating AI is automating RL environments Introducing General-Agent: A fully synthetic environment whose task corpus self-evolves and grows harder over time 4,504 tool-use tasks · 1,040 domains · 8,159 unique tools
#1153Florian Brand@XEOPHONmaking the tech that closed labs have open and giving it to everyone, one release after another :)
Prime Intellect@PrimeIntellectThe next step toward automating AI is automating RL environments Introducing General-Agent: A fully synthetic environment whose task corpus self-evolves and grows harder over time 4,504 tool-use tasks · 1,040 domains · 8,159 unique tools
#1186Johannes Hagemann@JOHANNES_HAGEgo from idea, to environment, to running 1000s of parallel, multi-hour agent episodes within a day.
open superintelligence stack
highly recommend checking out @mikasenghaas full blog post on the general agent environment release with all the details and experiments:
Johannes Hagemann@johannes_hagego from idea, to environment, to running 1000s of parallel, multi-hour agent episodes within a day. open superintelligence stack