INDEX

Explanations

uncommon characters or symbols

oai_token-act-pair · gpt-3.5-turbo

references to emotional states or reactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

aud

-0.78

 orchestr

-0.76

 swoop

-0.73

 steroids

-0.73

 imperson

-0.70

 hust

-0.69

 sacked

-0.68

 plunge

-0.68

 innocence

-0.68

 engagement

-0.68

POSITIVE LOGITS

ÂłÂłÂłÂłÂłÂłÂłÂł

1.87

ÂłÂłÂłÂł

1.63

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

1.60

Posted

1.56

Anyway

1.55

ÂłÂłÂł

1.51

ÂłÂł

1.41

posted

1.40

Anonymous

1.33

Conclusion

1.31

Activations Density 0.334%

uncommon characters or symbols

references to emotional states or reactions

No Comments

No Known Activations