INDEX

Explanations

words related to disapproval or criticism

oai_token-act-pair · gpt-3.5-turbo

negative connotations and criticisms of various subjects or topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ufact

-0.83

 hemor

-0.69

llular

-0.66

 reorgan

-0.65

 Farming

-0.63

 intest

-0.62

 raiding

-0.62

opez

-0.61

igsaw

-0.60

oufl

-0.60

POSITIVE LOGITS

 thereof

0.79

avorable

0.75

 regards

0.72

uous

0.72

 thereto

0.71

liest

0.71

 viewpoints

0.70

ably

0.69

soever

0.69

Supporters

0.68

Activations Density 0.101%

words related to disapproval or criticism

negative connotations and criticisms of various subjects or topics

No Comments

No Known Activations