INDEX

Explanations

Japanese names and terms

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or creators in the context of media production

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

adelphia

-0.80

roph

-0.74

sylv

-0.71

artisan

-0.71

undy

-0.66

stasy

-0.65

wild

-0.65

adr

-0.65

raped

-0.64

pes

-0.62

POSITIVE LOGITS

 Tanaka

1.42

 Yosh

1.41

 Nish

1.41

 Tsuk

1.39

 Makoto

1.39

 Suzuki

1.36

 Hiro

1.35

Miy

1.34

Nak

1.33

 Hirosh

1.33

Activations Density 0.128%

Japanese names and terms

references to specific individuals or creators in the context of media production

No Comments

No Known Activations