INDEX

Explanations

descriptions of research studies or legal affairs

oai_token-act-pair · gpt-3.5-turbo

key phrases related to ongoing events or studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 caut

-0.77

 externalToEVAOnly

-0.76

Secondly

-0.76

 Vaugh

-0.75

rapnel

-0.74

aviour

-0.72

adata

-0.71

nown

-0.71

ionage

-0.70

 superiors

-0.69

POSITIVE LOGITS

 âĢº

0.95

WASHINGTON

0.80

 Updated

0.79

0.73

HOU

0.72

 Posted

0.70

 Kavanaugh

0.68

 Jeep

0.68

 ....

0.67

Seattle

0.67

Activations Density 2.514%

descriptions of research studies or legal affairs

key phrases related to ongoing events or studies

No Comments

No Known Activations