A Taxonomy of Transcendence

• URL: http://arxiv.org/abs/2508.17669v1
• Date: Mon, 25 Aug 2025 05:05:00 GMT
• Title: A Taxonomy of Transcendence
• Authors: Natalie Abreu, Edwin Zhang, Eran Malach, Naomi Saphra,
• Abstract summary: We use a controlled setting to identify properties of the training data that lead a model to transcend the performance of its data sources.<n>We then introduce a knowledge graph-based setting in which simulated experts generate data based on their individual expertise.
• Score: 26.78660458573198
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Although language models are trained to mimic humans, the resulting systems display capabilities beyond the scope of any one person. To understand this phenomenon, we use a controlled setting to identify properties of the training data that lead a model to transcend the performance of its data sources. We build on previous work to outline three modes of transcendence, which we call skill denoising, skill selection, and skill generalization. We then introduce a knowledge graph-based setting in which simulated experts generate data based on their individual expertise. We highlight several aspects of data diversity that help to enable the model's transcendent capabilities. Additionally, our data generation setting offers a controlled testbed that we hope is valuable for future research in the area.

Related papers

• UNIFORM: Unifying Knowledge from Large-scale and Diverse Pre-trained Models [62.76435672183968]
  We introduce a novel framework, namely UNIFORM, for knowledge transfer from a diverse set of off-the-shelf models into one student model.<n>We propose a dedicated voting mechanism to capture the consensus of knowledge both at the logit level and at the feature level.<n>Experiments demonstrate that UNIFORM effectively enhances unsupervised object recognition performance compared to strong knowledge transfer baselines.
  arXiv  Detail & Related papers  (2025-08-27T00:56:11Z)
• Transcendence: Generative Models Can Outperform The Experts That Train Them [55.885802048647655]
  We study the phenomenon of transcendence: when a generative model achieves capabilities that surpass the abilities of the experts generating its data. We demonstrate transcendence by training an autoregressive transformer to play chess from game transcripts, and show that the trained model can sometimes achieve better performance than all players in the dataset.
  arXiv  Detail & Related papers  (2024-06-17T17:00:52Z)
• ORacle: Large Vision-Language Models for Knowledge-Guided Holistic OR Domain Modeling [41.30327565949726]
  We introduce ORacle, an advanced vision-language model designed for holistic OR domain modeling. It incorporates multi-view and temporal capabilities and can leverage external knowledge during inference, enabling it to adapt to previously unseen surgical scenarios. In rigorous testing, in scene graph generation, and downstream tasks on the 4D-OR dataset, ORacle not only demonstrates state-of-the-art performance but does so requiring less data than existing models.
  arXiv  Detail & Related papers  (2024-04-10T14:24:10Z)
• Capture the Flag: Uncovering Data Insights with Large Language Models [90.47038584812925]
  This study explores the potential of using Large Language Models (LLMs) to automate the discovery of insights in data. We propose a new evaluation methodology based on a "capture the flag" principle, measuring the ability of such models to recognize meaningful and pertinent information (flags) in a dataset.
  arXiv  Detail & Related papers  (2023-12-21T14:20:06Z)
• Fantastic Gains and Where to Find Them: On the Existence and Prospect of General Knowledge Transfer between Any Pretrained Model [74.62272538148245]
  We show that for arbitrary pairings of pretrained models, one model extracts significant data context unavailable in the other. We investigate if it is possible to transfer such "complementary" knowledge from one model to another without performance degradation.
  arXiv  Detail & Related papers  (2023-10-26T17:59:46Z)
• StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue Data [129.92449761766025]
  We propose a novel data collection methodology that synchronously synthesizes images and dialogues for visual instruction tuning. This approach harnesses the power of generative models, marrying the abilities of ChatGPT and text-to-image generative models. Our research includes comprehensive experiments conducted on various datasets.
  arXiv  Detail & Related papers  (2023-08-20T12:43:52Z)
• Towards A Unified Agent with Foundation Models [18.558328028366816]
  We investigate how to embed and leverage such abilities in Reinforcement Learning (RL) agents. We design a framework that uses language as the core reasoning tool, exploring how this enables an agent to tackle a series of fundamental RL challenges. We demonstrate substantial performance improvements over baselines in exploration efficiency and ability to reuse data from offline datasets.
  arXiv  Detail & Related papers  (2023-07-18T22:37:30Z)
• Data Quality in Imitation Learning [15.939363481618738]
  In offline learning for robotics, we simply lack internet scale data, and so high quality datasets are a necessity. This is especially true in imitation learning (IL), a sample efficient paradigm for robot learning using expert demonstrations. In this work, we take the first step toward formalizing data quality for imitation learning through the lens of distribution shift.
  arXiv  Detail & Related papers  (2023-06-04T18:48:32Z)
• An Empirical Investigation of Commonsense Self-Supervision with Knowledge Graphs [67.23285413610243]
  Self-supervision based on the information extracted from large knowledge graphs has been shown to improve the generalization of language models. We study the effect of knowledge sampling strategies and sizes that can be used to generate synthetic data for adapting language models.
  arXiv  Detail & Related papers  (2022-05-21T19:49:04Z)
• Interactive Weak Supervision: Learning Useful Heuristics for Data Labeling [19.24454872492008]
  Weak supervision offers a promising alternative for producing labeled datasets without ground truth labels. We develop the first framework for interactive weak supervision in which a method proposes iterations and learns from user feedback. Our experiments demonstrate that only a small number of feedback are needed to train models that achieve highly competitive test set performance.
  arXiv  Detail & Related papers  (2020-12-11T00:10:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.