INDEX

Explanations

technical terms related to different areas such as design, technology, and logistics

oai_token-act-pair · gpt-3.5-turbo

references to visual media or packaging details

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.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

oun

-0.99

 eleph

-0.99

 spir

-0.92

 gestation

-0.91

 ranged

-0.87

 metic

-0.87

 crumble

-0.85

 oppressed

-0.85

 scrambling

-0.85

 range

-0.85

POSITIVE LOGITS

License

1.88

Conclusion

1.84

Advertisements

1.77

Trivia

1.76

Other

1.75

Mis

1.75

Optional

1.72

References

1.72

Additional

1.70

Various

1.69

Activations Density 0.234%

technical terms related to different areas such as design, technology, and logistics

references to visual media or packaging details

No Comments

No Known Activations