INDEX

Explanations

words related to physical movement, specifically the action of swiping

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "sw" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

inelli

-0.86

onics

-0.83

endment

-0.80

OPA

-0.76

 Constantin

-0.74

viation

-0.74

oglu

-0.74

otaur

-0.73

Church

-0.71

exist

-0.71

POSITIVE LOGITS

sw

3.64

Sw

1.95

sw

1.77

Sw

1.75

 swat

1.65

 swipe

1.43

 swim

1.33

 swarm

1.28

SW

1.25

 swamp

1.22

Activations Density 0.010%

words related to physical movement, specifically the action of swiping

occurrences of the word "sw" in various contexts

No Comments

No Known Activations