Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models

• URL: http://arxiv.org/abs/2410.13826v2
• Date: Thu, 24 Oct 2024 17:27:22 GMT
• Title: Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
• Authors: Mazda Moayeri, Vidhisha Balachandran, Varun Chandrasekaran, Safoora Yousefi, Thomas Fel, Soheil Feizi, Besmira Nushi, Neel Joshi, Vibhav Vineet,
• Abstract summary: skill-wise performance is obscured when inspecting aggregate accuracy, under-utilizing the rich signal modern benchmarks contain. We propose an automatic approach to recover the underlying skills relevant for any evaluation instance, by way of inspecting model-generated rationales. Our skill-slices and framework open a new avenue in model evaluation, leveraging skill-specific analyses to unlock a more granular and actionable understanding of model capabilities.
• Score: 61.467781476005435
• License:
• Abstract: With models getting stronger, evaluations have grown more complex, testing multiple skills in one benchmark and even in the same instance at once. However, skill-wise performance is obscured when inspecting aggregate accuracy, under-utilizing the rich signal modern benchmarks contain. We propose an automatic approach to recover the underlying skills relevant for any evaluation instance, by way of inspecting model-generated rationales. After validating the relevance of rationale-parsed skills and inferring skills for $46$k instances over $12$ benchmarks, we observe many skills to be common across benchmarks, resulting in the curation of hundreds of skill-slices (i.e. sets of instances testing a common skill). Inspecting accuracy over these slices yields novel insights on model trade-offs: e.g., compared to GPT-4o and Claude 3.5 Sonnet, on average, Gemini 1.5 Pro is $18\%$ more accurate in "computing molar mass", but $19\%$ less accurate in "applying constitutional law", despite the overall accuracies of the three models differing by a mere $0.4\%$. Furthermore, we demonstrate the practical utility of our approach by showing that insights derived from skill slice analysis can generalize to held-out instances: when routing each instance to the model strongest on the relevant skills, we see a $3\%$ accuracy improvement over our $12$ dataset corpus. Our skill-slices and framework open a new avenue in model evaluation, leveraging skill-specific analyses to unlock a more granular and actionable understanding of model capabilities.

Related papers

• Self-rationalization improves LLM as a fine-grained judge [21.917301609125417]
  We introduce Self-Rationalization, an iterative process of improving the rationales for the judge models. Self-rationalization works by having the model generate multiple judgments with rationales for the same input. We show that our model learns to produce higher quality rationales, with a win rate of $62%$ on average compared to models just trained via SFT on rationale.
  arXiv  Detail & Related papers  (2024-10-07T21:05:53Z)
• Can Models Learn Skill Composition from Examples? [50.5142714905768]
  We evaluate the capacity of smaller models to learn compositional generalization from examples. We show that training on combinations of $k=2$ and $3$ skills results in noticeable improvements in the ability to compose texts. This study also suggests that incorporating skill-rich (potentially synthetic) text into training can substantially enhance the compositional capabilities of models.
  arXiv  Detail & Related papers  (2024-09-29T22:14:02Z)
• Learning Goal-Conditioned Representations for Language Reward Models [10.94845204766088]
  We propose training reward models (RMs) in a contrastive, $textitgoal-conditioned$ fashion. We show this way of training RM representations enables improved $textitsteerability$ because it allows us to evaluate the likelihood of an action achieving a particular goal-state. We additionally find that these representations can perform fine-grained control by conditioning on desired future goal-states.
  arXiv  Detail & Related papers  (2024-07-18T20:23:11Z)
• QualEval: Qualitative Evaluation for Model Improvement [82.73561470966658]
  We propose QualEval, which augments quantitative scalar metrics with automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights. We demonstrate that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15% points relative.
  arXiv  Detail & Related papers  (2023-11-06T00:21:44Z)
• Preserving Knowledge Invariance: Rethinking Robustness Evaluation of Open Information Extraction [50.62245481416744]
  We present the first benchmark that simulates the evaluation of open information extraction models in the real world. We design and annotate a large-scale testbed in which each example is a knowledge-invariant clique. By further elaborating the robustness metric, a model is judged to be robust if its performance is consistently accurate on the overall cliques.
  arXiv  Detail & Related papers  (2023-05-23T12:05:09Z)
• Feeding What You Need by Understanding What You Learned [54.400455868448695]
  Machine Reading (MRC) reveals the ability to understand a given text passage and answer questions based on it. Existing research works in MRC rely heavily on large-size models and corpus to improve the performance evaluated by metrics such as Exact Match. We argue that a deep understanding of model capabilities and data properties can help us feed a model with appropriate training data.
  arXiv  Detail & Related papers  (2022-03-05T14:15:59Z)
• Revisiting Model Stitching to Compare Neural Representations [8.331711958610347]
  We consider a "stitched model" formed by connecting the bottom-layers of $A$ to the top-layers of $B$, with a simple trainable layer between them. We show that good networks of the same architecture, but trained in very different ways, can be stitched to each other without drop in performance. We also give evidence for the intuition that "more is better" by showing that representations learnt with (1) more data, (2) bigger width, or (3) more training time can be "plugged in'' to weaker models to improve performance.
  arXiv  Detail & Related papers  (2021-06-14T18:05:10Z)
• Exploring Sparse Expert Models and Beyond [51.90860155810848]
  Mixture-of-Experts (MoE) models can achieve promising results with outrageous large amount of parameters but constant computation cost. We propose a simple method called expert prototyping that splits experts into different prototypes and applies $k$ top-$1$ routing. This strategy improves the model quality but maintains constant computational costs, and our further exploration on extremely large-scale models reflects that it is more effective in training larger models.
  arXiv  Detail & Related papers  (2021-05-31T16:12:44Z)

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.