INDEX

Explanations

phrases related to Metro locations or cities

oai_token-act-pair · gpt-3.5-turbo

references to different metropolitan areas

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

\\\\\\\\

-0.76

ifice

-0.75

ied

-0.75

ENGTH

-0.66

isted

-0.66

 testim

-0.66

 Rica

-0.65

 Kind

-0.65

acca

-0.64

ying

-0.64

POSITIVE LOGITS

plex

1.32

 Manila

1.07

lin

1.00

PC

0.95

jet

0.89

ropolitan

0.88

Stars

0.88

biology

0.88

bus

0.84

 Detroit

0.80

Activations Density 0.041%

phrases related to Metro locations or cities

references to different metropolitan areas

No Comments

No Known Activations