Genre
Transcendence: Generative Models Can Outperform The Experts That Train Them
Therefore, when trained on data generated by humans, we may not expect the artificial model to outperform the humans on their original objectives.
Guiding Large Language Models via Directional Stimulus Prompting
T5) to generate an auxiliary directional stimulus prompt for each input instance.
Unveiling User Satisfaction and Creator Productivity Trade-Offs in Recommendation Platforms
On User-Generated Content (UGC) platforms, recommendation algorithms significantly impact creators' motivation to produce content as they compete for
Reducing Transformer Key-Value Cache Size with Cross-Layer Attention William Brandon
Key-value (KV) caching plays an essential role in accelerating decoding for transformer-based autoregressive large language models (LLMs).
Disentangling Linear Quadratic Control with Untrusted ML Predictions Tongxin Li1 Hao Liu
We study the problem of online decision-making with predictions.
UDON: Universal Dynamic Online distillatioN for generic image representations
UDON's distillation approach is not only effective, but also very efficient, by sharing most model