INDEX

Explanations

terms related to moderate positions or individuals

oai_token-act-pair · gpt-3.5-turbo

the term "moderate" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

kefeller

-0.82

arium

-0.79

tyard

-0.78

çİĭ

-0.77

stals

-0.76

printed

-0.76

ifice

-0.73

ledge

-0.69

ilage

-0.68

yang

-0.67

POSITIVE LOGITS

 sized

1.06

erate

1.03

 minded

0.83

moderate

0.73

weights

0.71

 Islamists

0.71

 democr

0.71

(<

0.71

 moderate

0.70

 medi

0.69

Activations Density 0.052%

terms related to moderate positions or individuals

the term "moderate" in various contexts

No Comments

No Known Activations