INDEX

Explanations

phrases indicating a comparison or contrast

oai_token-act-pair · gpt-3.5-turbo

variations of the word "In" indicating various contexts or settings

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

IFrame

Link

Negative Logits

 unwelcome

-0.65

 reprinted

-0.63

 awaiting

-0.63

PDF

-0.62

 Amid

-0.62

ashington

-0.59

 antic

-0.57

 fugitive

-0.57

chenko

-0.57

 foothold

-0.56

POSITIVE LOGITS

entimes

1.02

 hindsight

1.00

Laughs

0.91

 somebody

0.88

Honestly

0.83

laughs

0.82

me

0.80

my

0.78

give

0.76

days

0.74

Activations Density 0.347%

phrases indicating a comparison or contrast

variations of the word "In" indicating various contexts or settings

No Known Activations