INDEX

Explanations

phrases related to moderation or moderate individuals

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "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 @ 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arium

-0.83

chy

-0.76

asper

-0.75

ynthesis

-0.74

STON

-0.73

æ©

-0.71

hon

-0.71

yang

-0.71

 Gleaming

-0.70

 Lust

-0.70

POSITIVE LOGITS

erate

1.00

 minded

0.87

 easing

0.80

 sized

0.79

yip

0.78

»Ĵ

0.77

 medi

0.77

 moderate

0.74

 leaning

0.73

moderate

0.71

Activations Density 0.013%

phrases related to moderation or moderate individuals

occurrences of the word "moderate" in various contexts

No Comments

No Known Activations