INDEX

Explanations

names, locations, and terms related to political and military contexts

oai_token-act-pair · gpt-3.5-turbo

names of people or entities associated with notable events or discussions

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

puter

-0.84

tub

-0.79

weed

-0.75

neau

-0.75

Madison

-0.74

linked

-0.70

ãĤ¤ãĥĪ

-0.70

loo

-0.68

wiki

-0.67

puff

-0.66

POSITIVE LOGITS

 Cron

0.94

yss

0.83

aucus

0.73

Pos

0.73

ezvous

0.73

 Pose

0.72

 Culture

0.72

zers

0.69

 Myth

0.69

rase

0.67

Activations Density 0.020%

names, locations, and terms related to political and military contexts

names of people or entities associated with notable events or discussions

No Comments

No Known Activations