INDEX

Explanations

phrases mentioning a specific keyword "ke" followed by a number, possibly indicating a keyword in a technical context or a code-related context

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "ke" or similar syllables in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

PsyNetMessage

-0.80

Downloadha

-0.78

 Skydragon

-0.72

western

-0.68

naire

-0.66

ãĤ¼ãĤ¦ãĤ¹

-0.66

ortium

-0.65

incial

-0.65

microsoft

-0.64

ãĥ´ãĤ¡

-0.64

POSITIVE LOGITS

jriwal

1.03

ez

0.91

eter

0.89

erk

0.89

ke

0.87

eling

0.86

llers

0.80

eto

0.80

lar

0.79

iency

0.78

Activations Density 0.004%

phrases mentioning a specific keyword "ke" followed by a number, possibly indicating a keyword in a technical context or a code-related context

the presence of the word "ke" or similar syllables in various contexts

No Comments

No Known Activations