Are We Learning Yet? A Meta Review of Evaluation Failures Across Machine Learning

By Thomas Liao. 2021

158 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Evaluation & Benchmarks

Citations in the corpus

Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text (2022, background): “In a recent survey of datasets and benchmarks in machine learning, Liao et al. (2021) point out that the lack of differentiation between tasks and datasets that aim to capture them can lead to harmful over-generalization.”
QLoRA: Efficient Finetuning of Quantized LLMs (2023, background): “Perhaps the largest is the problem of benchmark validity [36]—whether a benchmark truly tests what its name or description suggests is always at question, especially as we discover “shortcuts” to solve benchmarks that machine learning models sometimes exploit [22, 47].”
Data Governance in the Age of Large-Scale Data-Driven Language Technology (2022, methodology): “For LLMs in particular, the data used to train them are one further step removed from the task-specific models built from them, so the link between data and ML progress is even more abstracted [87, 116].”
MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts (2023, background): “High-quality evaluation datasets and benchmarks are a cornerstone for assessing the progress of machine learning models to solve real-world tasks Liao et al. (2021).”
Evaluation Gaps in Machine Learning Practice (2022, background): “The distinction between learner-centric and application-centric evaluations relates to the question of internal validity and external validity that is more commonly discussed in the social sciences than in ML (see, e.g., [123]) but also sometimes in ML [103].”
VisIT-Bench: A Benchmark for Vision-Language Instruction Following Inspired by Real-World Use (2023, background): “As a result, it is difficult to assess progress in this direction, which both reduces the field’s ability to identify true breakthroughs and increases vulnerability to potential pitfalls of evaluation that have hampered progress in other areas of machine learning [53, 54].”
The Fallacy of AI Functionality (2022, methodology): “[118] found that advances in machine learning often “evaporate under closer scrutiny or turn out to be less widely applicable than originally hoped.”