INDEX

Explanations

words related to music genres, particularly focusing on heavy metal and rock subgenres

oai_token-act-pair · gpt-3.5-turbo

musical genres, particularly various forms of metal and punk

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Ples

-0.77

 Shares

-0.73

 Mons

-0.71

 Serv

-0.70

pread

-0.69

 Volunte

-0.68

-0.66

andum

-0.64

 Housing

-0.64

urion

-0.63

POSITIVE LOGITS

 punk

1.05

 genre

1.05

 genres

1.05

punk

1.02

 music

0.99

 classics

0.99

 revival

0.99

core

0.97

 powerhouse

0.95

 legend

0.95

Activations Density 0.160%

words related to music genres, particularly focusing on heavy metal and rock subgenres

musical genres, particularly various forms of metal and punk

No Comments

No Known Activations