INDEX

Explanations

references to daily newsletters and email subscriptions

oai_token-act-pair · gpt-3.5-turbo

mentions of newsletters and subscription-related phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 unsolved

-0.67

 Dull

-0.64

wich

-0.61

gered

-0.60

 Revel

-0.60

ĪĴ

-0.59

 precincts

-0.58

itures

-0.56

 Woodward

-0.56

fired

-0.56

POSITIVE LOGITS

erity

0.65

DragonMagazine

0.63

ROR

0.61

nr

0.60

achine

0.59

oice

0.59

tymology

0.59

 Upgrade

0.58

å§«

0.57

dayName

0.56

Activations Density 0.020%

references to daily newsletters and email subscriptions

mentions of newsletters and subscription-related phrases

No Comments

No Known Activations