INDEX

Explanations

technical information related to computer programming, such as Java memory management and code snippets

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sculpt

-0.87

 covenant

-0.82

 retard

-0.77

 choking

-0.77

 revers

-0.76

 habit

-0.76

 brid

-0.76

 fleeing

-0.75

hyp

-0.75

 arche

-0.74

POSITIVE LOGITS

Advertisements

1.77

Conclusion

1.52

References

1.49

Loading

1.46

Contribut

1.45

Contact

1.44

Discussion

1.43

If

1.43

Trivia

1.42

Reply

1.40

Activations Density 1.323%

technical information related to computer programming, such as Java memory management and code snippets

No Comments

No Known Activations