INDEX

Explanations

references to architectural features and locations within a building

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_16/width_16k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.16.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

Ultimately

-0.41

 Ultimately

-0.39

melhor

-0.38

 ultimately

-0.36

blurRadius

-0.35

指针

-0.35

 Freed

-0.33

 spé

-0.33

 bookId

-0.33

 barb

-0.32

POSITIVE LOGITS

 hallway

0.91

 hallways

0.83

 foyer

0.73

 lobby

0.71

 downstairs

0.69

 Lobby

0.68

 lobbies

0.66

 stair

0.65

 vestibule

0.64

 pasillo

0.64

Activations Density 0.130%

references to architectural features and locations within a building

No Comments

No Known Activations