INDEX

Explanations

pieces of information related to attacks or incidents in various locations

oai_token-act-pair · gpt-3.5-turbo

references to tragic events or casualties

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

 emoji

-0.68

 cryptoc

-0.67

 freel

-0.67

 Maker

-0.66

xual

-0.65

UX

-0.65

 testers

-0.64

 Uber

-0.64

 cryptocurrency

-0.63

 talent

-0.63

POSITIVE LOGITS

Advertisements

1.46

Trivia

1.46

References

1.33

Discussion

1.31

Conclusion

1.28

CHAPTER

1.28

Notable

1.25

Detailed

1.24

PART

1.23

Appearances

1.21

Activations Density 0.524%

pieces of information related to attacks or incidents in various locations

references to tragic events or casualties

No Comments

No Known Activations