INDEX

Explanations

the word "anti" with varying levels of intensity

oai_token-act-pair · gpt-3.5-turbo

terms or phrases related to anti-government sentiments

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

mable

-0.89

ding

-0.74

sed

-0.73

nets

-0.71

wings

-0.71

lined

-0.70

manship

-0.70

come

-0.68

spring

-0.68

matically

-0.68

POSITIVE LOGITS

iso

0.92

qua

0.88

 Devi

0.88

anti

0.83

ucci

0.82

Ã±o

0.78

opsis

0.78

arius

0.75

qq

0.75

Å¡

0.74

Activations Density 0.028%

the word "anti" with varying levels of intensity

terms or phrases related to anti-government sentiments

No Comments

No Known Activations