INDEX

Explanations

mentions of specific geographical locations and events

oai_token-act-pair · gpt-3.5-turbo

references to specific locations and events

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

intel

-0.51

 fuzz

-0.50

×ķ

-0.50

ï¸ı

-0.50

oyal

-0.50

viz

-0.49

lihood

-0.49

 Hitman

-0.49

quo

-0.48

 accompl

-0.48

POSITIVE LOGITS

å£«

0.76

ãĤ¼ãĤ¦ãĤ¹

0.67

 Railway

0.61

stice

0.61

zona

0.60

 Stout

0.58

ukong

0.58

 Summit

0.57

 Wonderland

0.55

 ravaged

0.54

Activations Density 2.555%

mentions of specific geographical locations and events

references to specific locations and events

No Comments

No Known Activations