INDEX

Explanations

specific terms related to light and illumination

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

lighthouse, pituitary, Lightfoot, Poseidon, Lookout

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

Serbia

-0.44

 Serbia

-0.44

 Serbie

-0.43

 culo

-0.42

PreferredItem

-0.41

umán

-0.41

 Murcia

-0.40

Iraq

-0.39

Hauptartikel

-0.39

 scum

-0.38

POSITIVE LOGITS

 Lighthouse

2.36

 lighthouse

2.25

ighthouse

1.73

ligh

1.29

 ligh

1.11

 phare

0.83

 beacon

0.79

thouses

0.78

IntoConstraints

0.70

 faro

0.69

Activations Density 0.002%

specific terms related to light and illumination

lighthouse, pituitary, Lightfoot, Poseidon, Lookout

No Comments

No Known Activations