INDEX

Explanations

proper nouns, specifically names like "Jim"

oai_token-act-pair · gpt-3.5-turbo

the name "Jim" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

NetMessage

-0.85

DragonMagazine

-0.80

cffff

-0.80

BOOK

-0.80

Spoiler

-0.79

Constructed

-0.75

ylum

-0.74

CONCLUS

-0.70

Charges

-0.70

xual

-0.68

POSITIVE LOGITS

enez

1.33

mie

1.17

bo

0.88

 Yong

0.85

 Crow

0.85

 Beam

0.84

 Ross

0.80

bles

0.80

 Murphy

0.79

iny

0.78

Activations Density 0.009%

proper nouns, specifically names like "Jim"

the name "Jim" across various contexts

No Comments

No Known Activations