INDEX

Explanations

phrases related to specific entities or topics, possibly related to news, technology, and investigations

oai_token-act-pair · gpt-3.5-turbo

names and terms related to sensitive or private topics

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

CLOSE

-0.67

udeb

-0.64

ev

-0.62

share

-0.61

init

-0.59

ell

-0.54

hers

-0.54

cap

-0.53

ERO

-0.52

Van

-0.52

POSITIVE LOGITS

tainment

0.72

kefeller

0.69

itbart

0.68

terness

0.66

anyahu

0.65

osate

0.63

utenant

0.62

jriwal

0.62

merce

0.61

phabet

0.59

Activations Density 0.578%

phrases related to specific entities or topics, possibly related to news, technology, and investigations

names and terms related to sensitive or private topics

No Comments

No Known Activations