INDEX

Explanations

references to doctors and their titles

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

professional titles like "Dr." followed by a name, with high activation when the title and name are present.

oai_token-act-pair · claude-3-5-haiku-20241022 Triggered by @neilrathi

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 próxim

-0.56

 Eisenberg

-0.54

sæ

-0.53

ements

-0.52

ánló

-0.51

 Beal

-0.50

 Comet

-0.49

 Wain

-0.49

pagation

-0.48

 bezoek

-0.48

POSITIVE LOGITS

Dr

0.96

 Doctor

0.92

 doctor

0.91

Doctor

0.89

Dr

0.86

doctor

0.84

 DOCTOR

0.83

Drs

0.79

 docteur

0.75

 Doktor

0.73

Activations Density 0.057%

references to doctors and their titles

professional titles like "Dr." followed by a name, with high activation when the title and name are present.

No Comments

No Known Activations