INDEX

Explanations

technical terms and scholarly activities like studies, reports, and papers

oai_token-act-pair · gpt-3.5-turbo

references to research studies and related academic activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ĪĴ

-0.81

ãĤª

-0.71

yrics

-0.67

 trump

-0.64

 resume

-0.61

detail

-0.60

pite

-0.57

ertain

-0.57

animate

-0.56

 corners

-0.56

POSITIVE LOGITS

 himself

0.81

 herself

0.73

 anonymously

0.70

Ruk

0.69

 yourselves

0.69

 jointly

0.66

 Cheong

0.66

 Judd

0.62

etsk

0.61

 Yourself

0.61

Activations Density 0.371%

technical terms and scholarly activities like studies, reports, and papers

references to research studies and related academic activities

No Comments

No Known Activations