IC-1199GPT-2 next-token distributions contain correctable tail errors from the softmax bottleneck that degrade generation quality under low-entropy sampling, with basis-aware threshold sampling improving MAUVE across all four sizes

Matthew Finlayson, John Hewitt, Alexander Koller, Swabha Swayamdipta, Ashish Sabharwal

SourceClosing the Curious Case of Neural Text Degeneration

The paper demonstrates that GPT-2's low-rank softmax matrix causes probability overestimation for tokens outside the true support, leading to incoherent text under low-entropy (near-greedy) decoding. By exploiting the linear structure of the softmax matrix, the authors' basis-aware threshold (BAT) sampling selectively rejects high-probability but implausible tokens while retaining lower-probability but valid ones. Across all four GPT-2 sizes, BA-η outperforms standard η-sampling on MAUVE, with the gap largest for GPT-2-large. Human annotators also narrowly prefer BA-η generations over η-sampling generations.

Evidence
correlational
Key metric
MAUVE (OWT test): small η 85.0±1.4 vs ba-η 87.8±1.0; medium η 90.4±0.1 vs ba-η 92.2±0.6; large η 86.0±0.5 vs ba-η 88.4±0.5; xl η 87.1±1.2 vs ba-η 89.6±0.4. Human eval (ba-η≡0.002 vs η=0.002): 0.43/0.38/0.19; (ba-η≡0.024 vs η=0.024): 0.48/0.47/0.05
Caveat
Pilot study on 5000 prefixes; BAT does not pair well with nucleus sampling; for certain prefixes and low-entropy parameters, ba-ε accepts no tokens; linear program computation is expensive (amortized ~0.1 s/token); results limited to low-entropy generation settings
Model
GPT-2
Concepts
Failure mode
Datasets
OpenWebText / OpenWebText-10k [eval]
Methods
Nucleus Sampling [compared-to], MAUVE [eval]
Related work
Nucleus Sampling [context], Breaking the softmax bottleneck (Yang et al. 2018) [builds-on], Top-k sampling (Fan et al. 2018) [context]
Related findings
IC-1200
Extraction
automatic-extraction