Laminar introduced flow-1, a model trained with RL to find errors in agent traces. It matches GPT-6-sol in detection quality while being 23x cheaper, enabling analysis of every production agent trace without sampling.
The chart will appear after repeat observations. The current metric comes from the source.
12 source pointsReal observations only. History before source connection is not reconstructed.
The model enables effective analysis of agent traces, identifying hidden errors, which is crucial for ensuring system reliability.
Documents indicate that a police officer provided her Flock login to her husband, who is an ICE agent.
8 hours agoLessWrongThe article examines the impact of the λ parameter, which measures task parallelism, on the efficiency of AI swarms. The author analyzes how changes in λ affect computational resource scaling and AI development timelines.
17 hours agoLessWrongAnthropic released documents related to a copyright litigation case. The documents outline the company's research plans from 2021, including the development of a safe general AI assistant capable of programming tasks. The documents do not represent the company's official stance.
5 days agoLessWrongThe article discusses the existence of internal models that may not be publicly available and their impact on perceptions of AI capabilities and risks. It presents a graph showing a sharp increase in OpenAI's spending on AI agents, raising questions about its sustainability.
yesterdayHugging Face BlogFast private agents operating on-device using the LFM2.5-24B-A2B model are presented.
6 days agoLessWrongAs part of the MATS 9.1 extension program, mentored by Richard Ngo, this work explores applying the concept of cross-temporal incoherence to multistage games with observable actions. New equilibria defined through variants of the One Shot Deviation Principle (OSDP) are proposed, allowing analysis of situations where players reach inadequate equilibria with their future or counterfactual selves despite time-consistent discounting. The potential contributions of these variants to a theory of credible threats are also discussed.
3 days ago