INDEX

Explanations

strings of characters denoting server names

oai_token-act-pair · gpt-3.5-turbo

instances of a specific identifier or key term labeled "SN"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 funeral

-0.70

 potion

-0.64

 portrait

-0.63

 Canyon

-0.62

 drives

-0.61

 ........

-0.60

Wax

-0.60

uez

-0.59

Hel

-0.59

 procession

-0.58

POSITIVE LOGITS

SN

4.66

SN

1.79

SM

1.35

DN

1.34

SL

1.24

PN

1.22

NK

1.21

BN

1.20

NR

1.17

SW

1.16

Activations Density 0.007%

strings of characters denoting server names

instances of a specific identifier or key term labeled "SN"

No Comments

No Known Activations