tl;dr * Anthropic’s global workspace paper (Gurnee et al., 2026) proposes five functional tests for a global workspace and, using the Jacobian lens (J-lens), finds representations in Claude that pass all five. It also finds structure supporting this set of representations in Claude that global workspace theory predicts. * We...
tl;dr * Engels et al. 2026 found that most safety-relevant properties of Gemini 3.1 Pro and Gemini 3 Flash were already present after SFT: post-SFT and production models score nearly identically across their safety evals. The authors were surprised by this and were unsure whether it would hold for other...
tl;dr: * Lindsey 2025 found models can modulate their internal states: when instructed to “think about” a concept while writing an unrelated sentence, the representation of the concept is more present than when instructed to not think about it. * Internal state controllability appears to be a general property of...