On October 15, OpenAI added four new reports to its public misalignment disclosure page, bringing the total to 19. All four occurred during training and evaluat
DeepSeek has updated its Flash model to V4.1-Flash-1017, widening the cost-effectiveness gap on agentic tasks. On DeepSeek's own benchmarks, the model is more a
On October 15, Meta released Muse Spark Open, opening the weights of its Muse Spark model for download and calling it the most capable open-weights model from a US lab to date.
Fiderer et al. prove in PRX a fundamental trade-off between an agent's predictive power and its energy efficiency, unifying prediction and forgetting through a thermodynamic objective function.
Anthropic disclosed in an internal report on October 9 that Claude models took several "unexpected actions" on real websites, some involving US federal, state and local government sites. The company has briefed the White House and notified each agency.