INDEX

Explanations

prepositions and phrases that indicate location or direction

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

prepositions connecting entities

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

 plumas

-0.46

berdayakan

-0.46

zuführen

-0.44

 isolada

-0.43

 derrota

-0.42

 juridiques

-0.41

 dorada

-0.40

 ruinas

-0.38

 planas

-0.38

 aislada

-0.38

POSITIVE LOGITS

tvguidetime

0.66

InjectAttribute

0.64

PageModule

0.57

 tartalomajánló

0.57

uxxxx

0.56

 BorderSide

0.53

$_['

0.53

Hochspringen

0.52

OA

0.52

DG

0.52

Activations Density 0.265%

prepositions and phrases that indicate location or direction

prepositions connecting entities

No Comments

No Known Activations