Increasing robustness to spurious correlations using forgettable examples
… of pre-trained language models. In this paper, we first propose using example forgetting to
find … We consider three sentence pair classification tasks, namely natural language inference, …
find … We consider three sentence pair classification tasks, namely natural language inference, …
An empirical study on model-agnostic debiasing strategies for robust natural language inference
… on natural language inference (NLI) debiasing mainly targets at one or few known biases
while not necessarily making the models more robust. In this paper, we focus on the model-…
while not necessarily making the models more robust. In this paper, we focus on the model-…
Understanding catastrophic forgetting in language models via implicit inference
… to enable pretrained LLMs to follow natural language instructions. While instruction tuning
improves … and leverage transformer capabilities, providing robust and steerable methods for …
improves … and leverage transformer capabilities, providing robust and steerable methods for …
Unforgettable Generalization in Language Models
… Surprisingly, certain capabilities are robust to forgetting even after fine-tuning on random …
To qualify an example as forgotten, we require the model have a confidence of at most 0.6 in …
To qualify an example as forgotten, we require the model have a confidence of at most 0.6 in …
How should pre-trained language models be fine-tuned towards adversarial robustness?
… and natural language inference, under different attacks across various pre-trained … the
forgetting problem and achieve more robust models, we propose a novel approach named Robust …
forgetting problem and achieve more robust models, we propose a novel approach named Robust …
[HTML][HTML] Continual pre-training mitigates forgetting in language and vision
… nor about generic facts for language inference. Therefore, we … forgetting depends on the
input modality (natural language … , a naive fine-tuning approach is robust and does not show a …
input modality (natural language … , a naive fine-tuning approach is robust and does not show a …
Probing natural language inference models through semantic fragments
… As shown in the top part of Table 1, which we discuss later, we found that several strong
baselines … without optimal aggregate performance are often prone to catastrophic forgetting. …
baselines … without optimal aggregate performance are often prone to catastrophic forgetting. …
ANLIzing the adversarial natural language inference dataset
… Anyone working in the field will know that we are still far away from having models that can
perform NLI in a robust, … trained on later rounds don’t suffer from catastrophic forgetting. …
perform NLI in a robust, … trained on later rounds don’t suffer from catastrophic forgetting. …
e-snli: Natural language inference with natural language explanations
OM Camburu, T Rockt�schel…�- Advances in�…, 2018 - proceedings.neurips.cc
… at every timestep of the decoder, to avoid forgetting. … Since we collected 3 explanations for
each example in the … , we conclude that this measure is not reliable for our task, and we further …
each example in the … , we conclude that this measure is not reliable for our task, and we further …
Measuring forgetting of memorized training examples
… , we demonstrate a passive form of privacy amplification: examples used early in model
training may be more robust to privacy attacks. Our findings align with the theoretical findings of …
training may be more robust to privacy attacks. Our findings align with the theoretical findings of …