INDEX

Explanations

prepositions and phrases indicating relationships or connections between entities

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

preposition and noun

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

rungsseite

-0.50

felves

-0.48

andExpect

-0.47

arvi

-0.47

ConstraintMaker

-0.47

awtextra

-0.45

 cache

-0.42

áját

-0.42

 Exposition

-0.40

 cash

-0.40

POSITIVE LOGITS

цездатний

0.59

 '\\;'

0.50

principalTable

0.44

 الرياضيه

0.41

GOTREF

0.39

 informée

0.39

 Senhor

0.37

 linkovi

0.37

AsUp

0.37

AnimationsModule

0.37

Activations Density 0.115%

prepositions and phrases indicating relationships or connections between entities

preposition and noun

No Comments

No Known Activations