INDEX

Explanations

the word "strap."

oai_token-act-pair · gpt-3.5-turbo

references to straps or items used for securing or fastening

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

å¥³

-0.80

-0.72

DOM

-0.71

employment

-0.68

mberg

-0.68

Kenn

-0.67

cule

-0.66

administ

-0.65

ãĥīãĥ©ãĤ´ãĥ³

-0.63

official

-0.62

POSITIVE LOGITS

 strap

1.48

 straps

1.26

 buckle

1.04

 strapped

0.94

strap

0.93

 bracelet

0.89

 harness

0.84

onite

0.79

 implants

0.77

pipe

0.76

Activations Density 0.014%

the word "strap."

references to straps or items used for securing or fastening

No Comments

No Known Activations