INDEX

Explanations

percentages and numerical data

oai_token-act-pair · gpt-3.5-turbo

statistical metrics and numerical data comparisons

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

âĢ¦)

-0.64

too

-0.59

imaru

-0.59

LAR

-0.55

 Beyond

-0.55

achine

-0.54

ebin

-0.54

 enthusi

-0.52

)].

-0.52

 irresist

-0.50

POSITIVE LOGITS

 earlier

0.96

 previously

0.87

 previous

0.87

 usual

0.80

 originally

0.80

 prior

0.75

 customary

0.70

 normally

0.66

advert

0.65

 initially

0.63

Activations Density 0.164%

percentages and numerical data

statistical metrics and numerical data comparisons

No Comments

No Known Activations