INDEX

Explanations

numerical factors or values that are emphasized

oai_token-act-pair · gpt-3.5-turbo

instances of the word "factor" used in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

phis

-0.81

gem

-0.79

quartered

-0.77

escription

-0.77

ccording

-0.77

pit

-0.77

alach

-0.76

sterdam

-0.75

 Parables

-0.74

 Liberties

-0.73

POSITIVE LOGITS

 factor

1.05

factor

0.99

 factors

0.94

Factor

0.84

 Factor

0.83

 Factors

0.80

inity

0.78

uated

0.72

imate

0.71

uate

0.70

Activations Density 0.011%

numerical factors or values that are emphasized

instances of the word "factor" used in various contexts

No Comments

No Known Activations