Kürzlich aktualisierte Kanäle

Wählen, um nur diesen Kanal zu sehenAlle Kanäle →
Matthew Bermanheute +296AI Engineerheute +6Theo - t3․ggheuteThe Buildersheute +1a16zheute +2硅谷101heute +1Chris Williamsonheute20VC with Harry Stebbingsgestern +416Jane Streetgestern +119Y Combinatorgestern +1

Ergebnisse

5 Artikel
Sortieren
ThemaAI AlignmentPersonMatthew Berman
7 Folien15:13EN
Matthew Berman· 10. Mai 2025Absolute Zero: Training a Reasoner on Self-Generated Coding TasksMatthew Berman · Forward FutureMatthew Berman examines a 2025 paper in which a pretrained language model proposes and solves code tasks using verifiable feedback.Reinforcement LearningSynthetic dataReasoning & test-time compute+2
9 Folien24:08EN
Matthew Berman· 12. Dez. 2024What Apollo’s 2024 Scheming Tests Found in Language ModelsMatthew Berman · Forward FutureMatthew Berman reviews experiments on deceptive actions, concealment and sandbagging, with the study’s staged goals and sandbox limitations in view.AI SafetyEvalsAI Alignment+1
6 Folien13:39EN
Matthew Berman· 20. Nov. 2024What Interview-Based AI Agents Can Predict About PeopleMatthew Berman · Forward FutureMatthew Berman reviews a 2024 study using interviews with 1,052 participants to simulate survey answers and decisions.Multi-agent systemsSynthetic dataPsychology+2
2 Folien8:49EN
Matthew Berman· 19. Juli 2023Llama 2’s 2023 Release: Model Sizes, License and SafetyMatthew Berman · Forward FutureMatthew Berman reviews Meta’s Llama 2 paper, its commercial license, model variants and safety evaluations.Open Source & CommunityFoundation ModelsAI Alignment+1
3 Folien10:44EN
Matthew Berman· 9. Juli 2023Can AI Help Humans Align More Powerful AI?Matthew Berman · Forward FutureMatthew Berman examines OpenAI’s 2023 Superalignment proposal, its plans for AI-assisted oversight, and the unanswered question of who evaluates the evaluators.AI AlignmentAI SafetyHuman-in-the-Loop-Validierung+2