INDEX

Explanations

proper nouns or acronyms related to business or organizations

oai_token-act-pair · gpt-3.5-turbo

mentions of corporations or company names

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

 hypot

-0.63

 doub

-0.61

 diplom

-0.60

 Pascal

-0.58

 whole

-0.56

 dwarves

-0.56

 Ãľ

-0.56

 entire

-0.56

ard

-0.55

 surve

-0.55

POSITIVE LOGITS

guyen

1.31

uclear

1.21

uggets

1.20

ominated

1.12

ihil

1.12

OVA

1.11

urse

1.09

onsense

1.06

elson

1.04

aughty

1.04

Activations Density 0.038%

proper nouns or acronyms related to business or organizations

mentions of corporations or company names

No Comments

No Known Activations