INDEX

Explanations

contracted words with an apostrophe in them

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific symbol or character

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

 silhou

-0.62

 decomp

-0.59

 Mush

-0.58

 Stuff

-0.58

 Laurent

-0.56

 contact

-0.55

 JPEG

-0.55

 Mobil

-0.55

 neut

-0.54

Yor

-0.54

POSITIVE LOGITS

0.89

ates

0.86

uble

0.85

ese

0.85

tal

0.84

ski

0.83

tarian

0.80

al

0.79

meric

0.78

tarians

0.76

Activations Density 0.271%

contracted words with an apostrophe in them

occurrences of a specific symbol or character

No Comments

No Known Activations