INDEX

Explanations

the word "Wild."

oai_token-act-pair · gpt-3.5-turbo

references to the term "Wild" associated with various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

uyomi

-0.89

akeru

-0.82

uters

-0.81

 mathemat

-0.79

berus

-0.77

acists

-0.76

ayers

-0.74

ä»£

-0.73

aminer

-0.73

adian

-0.72

POSITIVE LOGITS

erness

1.04

lings

0.99

fire

0.91

fires

0.90

flower

0.90

mother

0.85

tro

0.85

enstein

0.82

Tang

0.80

cat

0.79

Activations Density 0.005%

the word "Wild."

references to the term "Wild" associated with various contexts

No Comments

No Known Activations