INDEX

Explanations

non-alphanumeric characters and text formatting markup

oai_token-act-pair · gpt-3.5-turbo

symbols or formatting that indicate a transition or change

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 powering

-0.58

 telecommunications

-0.57

 basket

-0.56

 laying

-0.56

 baskets

-0.55

swe

-0.55

 broadcasting

-0.55

 routed

-0.53

ilant

-0.53

perty

-0.52

POSITIVE LOGITS

 Conclusion

0.82

 References

0.76

 Continued

0.73

ĻĤ

0.72

________________

0.71

ibliography

0.70

 Quote

0.66

laughs

0.66

Â¶

0.66

 Interview

0.63

Activations Density 0.574%

non-alphanumeric characters and text formatting markup

symbols or formatting that indicate a transition or change

No Comments

No Known Activations