INDEX

Explanations

mentions of writing or coding-related activities

oai_token-act-pair · gpt-3.5-turbo

instances of writing and related activities

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

osate

-0.93

terness

-0.89

yip

-0.88

-0.85

ESA

-0.79

ordering

-0.79

ridor

-0.78

ĸļ

-0.77

pling

-0.74

ENTION

-0.70

POSITIVE LOGITS

 professionally

1.12

 weddings

1.02

 comics

0.96

 documentaries

0.95

 podcasts

0.91

 novels

0.91

 music

0.88

 videog

0.86

 movies

0.85

 mysteries

0.83

Activations Density 0.363%

mentions of writing or coding-related activities

instances of writing and related activities

No Comments

No Known Activations