INDEX

Explanations

Roman numeral and letter combinations

oai_token-act-pair · gpt-3.5-turbo

references to musical terms or entities, including songs, artists, and music-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Vaugh

-0.74

 toile

-0.74

redo

-0.66

CBI

-0.64

 Notting

-0.63

 fate

-0.63

mot

-0.62

 questioning

-0.62

 vulnerability

-0.60

 abdom

-0.60

POSITIVE LOGITS

intosh

0.93

arella

0.92

htar

0.92

ufact

0.83

ouf

0.82

agara

0.80

nih

0.80

gage

0.78

ogi

0.78

ONEY

0.78

Activations Density 0.600%

Roman numeral and letter combinations

references to musical terms or entities, including songs, artists, and music-related concepts

No Comments

No Known Activations