INDEX

Explanations

phrases indicating location, placement, and prepositions related to spatial relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 chi̍t

-0.68

 referrerpolicy

-0.64

 maroc

-0.58

swire

-0.57

MigrationBuilder

-0.57

 carats

-0.57

openzeppelin

-0.55

WireFormatLite

-0.55

Rena

-0.55

 Moslem

-0.54

POSITIVE LOGITS

 importantly

0.69

何より

0.57

 many

0.54

 lẫn

0.51

Через

0.51

 וב

0.51

 some

0.50

 much

0.50

 تضيفلها

0.50

Σε

0.50

Activations Density 0.232%

phrases indicating location, placement, and prepositions related to spatial relationships

No Comments

No Known Activations