INDEX

Explanations

terms related to positioning or placement in various contexts

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

how something is positioned

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

GPL

-0.42

 esternos

-0.40

Ծանոթ

-0.37

 ddelweddau

-0.36

Ohne

-0.36

isFull

-0.36

zanas

-0.35

 penas

-0.35

 Ohne

-0.35

igshid

-0.35

POSITIVE LOGITS

 Positioning

0.85

Positioning

0.85

 positioning

0.85

 positioned

0.82

positioning

0.73

 POSITION

0.71

Positioned

0.71

 posicion

0.71

 reposition

0.67

 estratég

0.65

Activations Density 0.010%

terms related to positioning or placement in various contexts

how something is positioned

No Comments

No Known Activations