INDEX

Explanations

capitalized abbreviations followed by numbers

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "IS" in various contexts

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

ruciating

-0.80

 Mercury

-0.69

 Meridian

-0.67

¥ŀ

-0.66

inges

-0.65

¥µ

-0.64

ished

-0.63

è¦ļéĨĴ

-0.63

 Thousand

-0.62

 Moor

-0.62

POSITIVE LOGITS

IS

1.27

BN

1.18

SON

1.07

FI

1.03

HER

1.01

IL

1.00

TY

1.00

KI

1.00

TRY

0.99

SI

0.99

Activations Density 0.009%

capitalized abbreviations followed by numbers

occurrences of the term "IS" in various contexts

No Comments

No Known Activations