INDEX

Explanations

names of specific locations or regions

oai_token-act-pair · gpt-3.5-turbo

geographic locations or proper nouns related to specific regions or entities

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

Phys

-0.83

bleacher

-0.81

ãĤ´ãĥ³

-0.79

 Wally

-0.79

ĸļ

-0.79

 Boards

-0.75

Reviewer

-0.74

 Grateful

-0.74

worms

-0.73

 Cthulhu

-0.71

POSITIVE LOGITS

 annexed

0.86

 unpop

0.83

 citiz

0.83

 govern

0.82

 nationalist

0.80

atan

0.78

 Hague

0.78

 dictatorship

0.78

 bloc

0.77

 confiscated

0.77

Activations Density 0.392%

names of specific locations or regions

geographic locations or proper nouns related to specific regions or entities

No Comments

No Known Activations