INDEX

Explanations

keywords related to the action of turning something off

oai_token-act-pair · gpt-3.5-turbo

phrases relating to deactivation or turning off something

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

Gall

-0.72

ONEY

-0.68

geries

-0.65

ysical

-0.63

oried

-0.63

 Reloaded

-0.62

gered

-0.61

ISM

-0.61

ophile

-0.60

ĸļ

-0.60

POSITIVE LOGITS

ment

0.82

ments

0.75

screen

0.73

 coasts

0.71

shoot

0.68

loading

0.67

 coast

0.66

 leash

0.65

sob

0.65

 caps

0.64

Activations Density 0.024%

keywords related to the action of turning something off

phrases relating to deactivation or turning off something

No Comments

No Known Activations