INDEX

Explanations

references to medical professionals or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

-------------</

-0.52

 ſta

-0.50

 arşivlendi

-0.48

ſelf

-0.47

AndEndTag

-0.47

☚

-0.45

stützung

-0.44

ualitas

-0.44

identité

-0.44

 Nachweis

-0.44

POSITIVE LOGITS

 Professor

0.71

Professor

0.65

Senator

0.54

 Aunt

0.54

Aunt

0.52

 Senator

0.51

 Rabbi

0.50

 Reverend

0.50

Dr

0.50

 Lieutenant

0.50

Activations Density 0.311%

references to medical professionals or titles

No Comments

No Known Activations