INDEX

Explanations

words related to museums or art institutions

oai_token-act-pair · gpt-3.5-turbo

references to museums

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

nir

-0.78

Leary

-0.70

LAN

-0.68

 Wast

-0.66

comes

-0.64

lees

-0.63

ornia

-0.63

OTH

-0.62

shaw

-0.60

---------

-0.60

POSITIVE LOGITS

useum

1.08

ĸļ

1.02

 curator

1.01

 museum

0.99

 museums

0.93

arium

0.88

 exhibits

0.86

 exhibit

0.82

 galleries

0.81

artifacts

0.81

Activations Density 0.019%

words related to museums or art institutions

references to museums

No Comments

No Known Activations