INDEX

Explanations

strong negative connotations

oai_token-act-pair · gpt-3.5-turbo

instances or expressions of the word "dumb" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

AUT

-0.74

aunder

-0.70

thus

-0.69

________________________________________________________________

-0.69

cially

-0.66

tes

-0.66

aukee

-0.66

Downloadha

-0.65

TOR

-0.65

OHN

-0.65

POSITIVE LOGITS

 dumb

1.15

bell

0.90

founded

0.88

wallet

0.87

 Dumb

0.84

glers

0.83

 stupid

0.82

gest

0.80

 reflex

0.78

 Stupid

0.78

Activations Density 0.002%

strong negative connotations

instances or expressions of the word "dumb" in various contexts

No Comments

No Known Activations