INDEX

Explanations

mentions of the concept of 'mind'

oai_token-act-pair · gpt-3.5-turbo

references to the concept of the mind

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

reek

-0.71

EStreamFrame

-0.68

 audible

-0.63

DIR

-0.62

interstitial

-0.60

 Zombies

-0.59

aukee

-0.58

Pie

-0.57

 Giuliani

-0.57

ded

-0.56

POSITIVE LOGITS

fulness

0.99

mind

0.99

Mind

0.82

bender

0.79

 mind

0.78

plex

0.76

ful

0.76

lock

0.75

erves

0.75

fully

0.74

Activations Density 0.017%

mentions of the concept of 'mind'

references to the concept of the mind

No Comments

No Known Activations