INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

specific geographical locations and names related to a certain area or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Nevada

-0.74

Ana

-0.70

 Rove

-0.68

 Liber

-0.67

 Juliet

-0.67

 Photographer

-0.65

aceae

-0.65

RH

-0.64

RH

-0.64

eers

-0.63

POSITIVE LOGITS

onge

1.25

 streetcar

0.93

TTC

0.92

 subway

0.88

 Mississ

0.84

orough

0.83

usable

0.81

 Subway

0.81

uffer

0.80

ppard

0.80

Activations Density 0.017%

locations or places

specific geographical locations and names related to a certain area or events

No Comments

No Known Activations