INDEX

Explanations

phrases related to giving permission or making something operational

oai_token-act-pair · gpt-3.5-turbo

instances of the word "enable" and its variations, indicating a focus on facilitation or empowerment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

apest

-0.84

boy

-0.78

ioxide

-0.72

ella

-0.66

iform

-0.66

baby

-0.66

alian

-0.66

anne

-0.66

 Irving

-0.65

ĳå£«

-0.65

POSITIVE LOGITS

ment

0.88

ments

0.84

MENTS

0.73

untarily

0.71

 interoper

0.70

ances

0.67

us

0.66

mentation

0.66

issance

0.64

MENT

0.63

Activations Density 0.031%

phrases related to giving permission or making something operational

instances of the word "enable" and its variations, indicating a focus on facilitation or empowerment

No Comments

No Known Activations