INDEX

Explanations

the word "motion."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "motion."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aez

-0.75

rosso

-0.69

arding

-0.68

ythm

-0.67

addy

-0.65

obin

-0.65

dds

-0.64

============

-0.64

angers

-0.63

nect

-0.62

POSITIVE LOGITS

 motion

1.32

motion

1.27

 motions

1.22

 Motion

1.18

Motion

1.11

Animation

0.85

 sickness

0.84

entary

0.80

 blur

0.78

ngth

0.78

Activations Density 0.012%

the word "motion."

occurrences of the word "motion."

No Comments

No Known Activations