INDEX

Explanations

phrases related to privacy or privilege

oai_token-act-pair · gpt-3.5-turbo

terms related to privacy and privilege

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

gaard

-0.68

 Ryder

-0.63

 Benedict

-0.62

GER

-0.62

 Modest

-0.60

 Glob

-0.58

Gen

-0.58

 Frankenstein

-0.56

 Danielle

-0.56

LER

-0.56

POSITIVE LOGITS

ilege

1.59

ileged

1.51

ately

1.02

ilage

1.00

opoly

0.96

opol

0.95

folios

0.93

ropri

0.89

uded

0.86

omore

0.86

Activations Density 0.055%

phrases related to privacy or privilege

terms related to privacy and privilege

No Comments

No Known Activations