brittle lily
An error occurred when submitting the following prompt, and interestingly, the error may occur stochastically and not occur if executed after a certain amount of time. It is possible that the error is due to a checking mechanism for the generated text in JSON format, as the prompt is designed to generate JSON-formatted text.
prompt:
Given title and summary of an arXiv paper, classify it into one of the following categories and output the probabilities:
## Categories
- text_generation
- speech_recognition
- speech_synthesis
- music_generation
- image_generation
- image_recognition
- object_detection
- video_generation
- video_analysis
- nerf
- others
## Input / Output
input: {title: title, summary: summary}
output: {reason: reason, category: category_probabilities}
## Test 1
input: {title: Scaling Vision Transformers to 22 Billion Parameters, summary: 'The scaling of Transformers has driven breakthrough capabilities for language models. At present, the largest large language models (LLMs) contain upwards of 100B parameters. Vision Transformers (ViT) have introduced the same architecture to image and video modelling, but these have not yet been successfully scaled to nearly the same degree; the largest dense ViT contains 4B parameters (Chen et al., 2022). We present a recipe for highly efficient and stable training of a 22B-parameter ViT (ViT-22B) and perform a wide variety of experiments on the resulting model. When evaluated on downstream tasks (often with a lightweight linear model on frozen features), ViT-22B demonstrates increasing performance with scale. We further observe other interesting benefits of scale, including an improved tradeoff between fairness and performance, state-of-the-art alignment to human visual perception in terms of shape/texture bias, and improved robustness. ViT-22B demonstrates the potential for LLM-like scaling in vision, and provides key steps towards getting there.}
output: {reason: The paper describes a highly efficient and stable training recipe for a 22B-parameter Vision Transformer (ViT-22B) and presents a wide range of experiments on the model, demonstrating its increasing performance with scale on downstream tasks. The focus of the paper is on scaling up the architecture to improve image recognition performance, making it a fit for the image recognition genre., category: {image_generation: 0.82, image_recognition: 0.12, others: 0.06}}
## Test 2
input: {title: A Song of Ice and Fire: Analyzing Textual Autotelic Agents in ScienceWorld, summary: Building open-ended agents that can autonomously discover a diversity of behaviours is one of the long-standing goals of artificial intelligence. This challenge can be studied in the framework of autotelic RL agents, i.e. agents that learn by selecting and pursuing their own goals, self-organizing a learning curriculum. Recent work identified language has a key dimension of autotelic learning, in particular because it enables abstract goal sampling and guidance from social peers for hindsight relabelling. Within this perspective, we study the following open scientific questions: What is the impact of hindsight feedback from a social peer (e.g. selective vs. exhaustive)? How can the agent learn from very rare language goal examples in its experience replay? How can multiple forms of exploration be combined, and take advantage of easier goals as stepping stones to reach harder ones? To address these questions, we use ScienceWorld, a textual environment with rich abstract and combinatorial physics. We show the importance of selectivity from the social peers feedback; that experience replay needs to over-sample examples of rare goals; and that following self-generated goal sequences where the agents competence is intermediate leads to significant improvements in final performance.}
output: