Publications (6)
Covert Malicious Finetuning: Challenges in Safeguarding LLM Adaptation
Danny Halawi, Alexander Wei, Eric Wallace +3
Black-box finetuning is an emerging interface for adapting state-of-the-art language models to user needs. However, such access may also let malicious actors undermine model safety…
Eliciting Latent Predictions from Transformers with the Tuned Lens
Nora Belrose, Igor Ostrovsky, Lev McKinney +5
We analyze transformers from the perspective of iterative inference, seeking to understand how model predictions are refined layer by layer. To do so, we train an affine probe for…
ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilities
Ezra Karger, Houtan Bastani, Chen Yueh-Han +4
Forecasts of future events are essential inputs into informed decision-making. Machine learning (ML) systems have the potential to deliver forecasts at scale, but there is no frame…
Dominion: A New Frontier for AI Research
Danny Halawi, Aron Sarmasi, Siena Saltzen +1
In recent years, machine learning approaches have made dramatic advances, reaching superhuman performance in Go, Atari, and poker variants. These games, and others before them, hav…
Approaching Human-Level Forecasting with Language Models
Danny Halawi, Fred Zhang, Chen Yueh-Han +1
Forecasting future events is important for policy and decision making. In this work, we study whether language models (LMs) can forecast at the level of competitive human forecaste…
Overthinking the Truth: Understanding how Language Models Process False Demonstrations
Danny Halawi, Jean-Stanislas Denain, Jacob Steinhardt
Modern language models can imitate complex patterns through few-shot learning, enabling them to complete challenging tasks without fine-tuning. However, imitation can also lead mod…