INDEX

Explanations

activities related to scrolling or movements within a document or interface

oai_token-act-pair · gpt-3.5-turbo

instances of the word "scroll" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

stood

-0.90

ilities

-0.71

Sov

-0.70

noxious

-0.70

anol

-0.67

jured

-0.66

rolet

-0.65

ãĤº

-0.64

tained

-0.64

responsible

-0.63

POSITIVE LOGITS

 scrolling

0.98

 scroll

0.96

 Scroll

0.96

 scrolls

0.91

scroll

0.89

antry

0.83

wheel

0.81

itzer

0.79

bars

0.78

chemy

0.77

Activations Density 0.016%

activities related to scrolling or movements within a document or interface

instances of the word "scroll" and its variations

No Comments

No Known Activations