INDEX

Explanations

words related to repeaters or repetitive actions

oai_token-act-pair · gpt-3.5-turbo

various forms of the word "repeat" or "repeal."

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

 Loans

-0.74

doms

-0.74

 Transparency

-0.73

 Fifth

-0.72

 GOODMAN

-0.71

heid

-0.68

 Accountability

-0.66

 Bahrain

-0.66

 Capitalism

-0.65

Io

-0.65

POSITIVE LOGITS

 repe

1.22

lling

1.00

lled

0.96

ptions

0.93

ple

0.93

naissance

0.91

plet

0.91

uel

0.90

bles

0.88

fect

0.87

Activations Density 0.012%

words related to repeaters or repetitive actions

various forms of the word "repeat" or "repeal."

No Comments

No Known Activations