INDEX

Explanations

arguments or opinions presented in text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "argue" and its variations, indicating discussions of opinions or beliefs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

beam

-0.70

fortunately

-0.68

haar

-0.66

cy

-0.65

ammy

-0.65

nown

-0.65

eco

-0.65

incarn

-0.64

mental

-0.64

cler

-0.63

POSITIVE LOGITS

 arguments

0.98

 arguing

0.94

 argue

0.90

 argued

0.89

 argues

0.84

 persu

0.81

 argument

0.79

 against

0.77

 vehemently

0.77

uca

0.76

Activations Density 0.030%

arguments or opinions presented in text

instances of the word "argue" and its variations, indicating discussions of opinions or beliefs

No Comments

No Known Activations