INDEX

Explanations

phrases related to physical locations or settings

oai_token-act-pair · gpt-3.5-turbo

references to "the rest of the" followed by various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Pwr

-0.68

 Missile

-0.67

ucci

-0.66

/+

-0.66

OPA

-0.65

Magikarp

-0.64

uca

-0.63

EStream

-0.62

ĸļ

-0.62

POSITIVE LOGITS

lishes

0.90

ulty

0.76

ordinary

0.72

course

0.72

emetery

0.72

country

0.72

abase

0.71

ractions

0.71

atform

0.68

iates

0.68

Activations Density 0.065%

phrases related to physical locations or settings

references to "the rest of the" followed by various subjects

No Comments

No Known Activations