INDEX

Explanations

references to doors

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

door and door-related phrases

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 doméstico

-0.38

 mínimos

-0.38

 normalidad

-0.36

 MainAxisSize

-0.36

 élève

-0.34

 orgánico

-0.34

 redacción

-0.34

 desierto

-0.34

 adquis

-0.34

 sagrado

-0.34

POSITIVE LOGITS

 door

1.11

 Door

1.06

Door

0.98

door

0.95

 DOOR

0.89

 through

0.88

 THROUGH

0.82

 przez

0.81

 Through

0.81

 durch

0.81

Activations Density 0.000%

references to doors

door and door-related phrases

No Comments

No Known Activations