INDEX

Explanations

phrases related to authority or official positions

oai_token-act-pair · gpt-3.5-turbo

terms related to ownership or value

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Skydragon

-0.69

 conviction

-0.61

eng

-0.60

 srfAttach

-0.55

 respons

-0.55

 Newton

-0.54

OVER

-0.52

abit

-0.52

RPG

-0.52

 vest

-0.52

POSITIVE LOGITS

aneous

0.89

theless

0.79

aneously

0.76

tenance

0.76

phrine

0.73

\":

0.72

areth

0.70

terday

0.69

'';

0.68

andum

0.66

Activations Density 0.253%

phrases related to authority or official positions

terms related to ownership or value

No Comments

No Known Activations