INDEX

Explanations

references to zombies and vampires

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

zombies and vampires

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

draw

-0.39

Draw

-0.37

ặt

-0.37

 draw

-0.36

 Infórmanos

-0.35

 Place

-0.34

plaats

-0.34

 Draw

-0.34

 GetEnumerator

-0.33

ностях

-0.33

POSITIVE LOGITS

 zombie

1.27

 Zombie

1.23

Zombie

1.19

zombie

1.19

 zombies

1.05

 Zombies

0.98

Zombies

0.87

ombies

0.86

ombie

0.77

🧟

0.76

Activations Density 0.005%

references to zombies and vampires

zombies and vampires

No Comments

No Known Activations