INDEX

Explanations

words related to deviation, dissension, or diversion from a norm or path

oai_token-act-pair · gpt-3.5-turbo

forms of the verb "mediate" and related words

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

=-=-=-=-=-=-=-=-

-0.70

shop

-0.66

micro

-0.65

tw

-0.61

 planner

-0.60

ington

-0.60

sonian

-0.60

 redes

-0.60

speed

-0.59

ingen

-0.59

POSITIVE LOGITS

iate

1.41

iates

1.23

iated

1.15

iating

1.13

iation

0.94

iations

0.83

lyak

0.75

rarily

0.72

uations

0.71

iator

0.70

Activations Density 0.013%

words related to deviation, dissension, or diversion from a norm or path

forms of the verb "mediate" and related words

No Comments

No Known Activations