INDEX

Explanations

instances where someone is personally involved or mentioned

oai_token-act-pair · gpt-3.5-turbo

the word "personally."

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

iens

-0.85

 Ends

-0.73

 Tycoon

-0.70

 Gates

-0.68

 Landing

-0.66

 Stall

-0.66

 Weaver

-0.65

 Conclusion

-0.65

xual

-0.64

 Faster

-0.63

POSITIVE LOGITS

 identifiable

1.24

ised

0.88

 benefited

0.84

 intervened

0.83

 invested

0.83

 speaking

0.82

 minded

0.80

 opposed

0.77

 insulted

0.77

 advising

0.76

Activations Density 0.012%

instances where someone is personally involved or mentioned

the word "personally."

No Comments

No Known Activations