INDEX

Explanations

words related to power and authority

oai_token-act-pair · gpt-3.5-turbo

terms related to severity or seriousness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

adr

-0.69

spring

-0.63

INTER

-0.63

HOU

-0.62

CHR

-0.62

onement

-0.61

 Cosponsors

-0.59

ISO

-0.59

 Mori

-0.58

otle

-0.58

POSITIVE LOGITS

vere

1.37

ignt

0.93

ndum

0.91

tenance

0.87

erate

0.79

ignty

0.78

mber

0.76

nda

0.75

ll

0.75

ble

0.75

Activations Density 0.005%

words related to power and authority

terms related to severity or seriousness

No Comments

No Known Activations