INDEX

Explanations

adjectives related to negative impact, such as 'hurt'

oai_token-act-pair · gpt-3.5-turbo

instances of the word "hurt" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

DragonMagazine

-0.69

vironment

-0.68

iasco

-0.65

au

-0.64

uties

-0.64

clerosis

-0.64

guyen

-0.63

aut

-0.63

aer

-0.62

 pedigree

-0.62

POSITIVE LOGITS

ful

0.98

 hurt

0.91

onies

0.87

 hurts

0.81

headed

0.80

 hurting

0.78

igue

0.78

 feelings

0.78

fully

0.76

lehem

0.75

Activations Density 0.011%

adjectives related to negative impact, such as 'hurt'

instances of the word "hurt" in various contexts

No Comments

No Known Activations