INDEX

Explanations

phrases related to removing or eliminating something

oai_token-act-pair · gpt-3.5-turbo

phrases about eliminating unwanted elements or experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 McDonnell

-0.63

 Luxembourg

-0.62

anwhile

-0.61

Leaks

-0.58

foreseen

-0.57

 McKenzie

-0.57

KEN

-0.57

 Belg

-0.56

 Belief

-0.55

 Johns

-0.54

POSITIVE LOGITS

icide

0.89

ged

0.85

icides

0.80

urga

0.79

ta

0.78

ients

0.77

uctions

0.76

estone

0.74

itary

0.73

ptin

0.73

Activations Density 0.021%

phrases related to removing or eliminating something

phrases about eliminating unwanted elements or experiences

No Comments

No Known Activations