INDEX

Explanations

words related to literature or author names

oai_token-act-pair · gpt-3.5-turbo

the presence of significant numerical data or categorizations related to people or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 faire

-0.75

uca

-0.73

regon

-0.73

awa

-0.70

 princ

-0.70

 footing

-0.68

 scrap

-0.67

swer

-0.66

xual

-0.65

 opposite

-0.62

POSITIVE LOGITS

è¦ļéĨĴ

0.87

Snow

0.80

Newsletter

0.80

Narr

0.79

Official

0.79

Disclaimer

0.76

ãĥŁ

0.75

Tags

0.74

dyl

0.74

Recommended

0.73

Activations Density 0.204%

words related to literature or author names

the presence of significant numerical data or categorizations related to people or concepts

No Comments

No Known Activations