INDEX

Explanations

the word "like" followed by different scenarios or preferences expressed in the text

oai_token-act-pair · gpt-3.5-turbo

expressions of dislike or aversion

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

iggins

-0.84

rontal

-0.83

utical

-0.81

yrinth

-0.81

itech

-0.79

owa

-0.76

nown

-0.75

verty

-0.74

oided

-0.73

olon

-0.73

POSITIVE LOGITS

 anymore

0.96

 anybody

0.88

nor

0.87

 anything

0.80

 anyone

0.79

any

0.76

ably

0.75

 surprises

0.71

 censorship

0.67

 bullies

0.66

Activations Density 0.048%

the word "like" followed by different scenarios or preferences expressed in the text

expressions of dislike or aversion

No Comments

No Known Activations