INDEX

Explanations

the term "grand" and its variations, indicating a focus on grandeur or significance

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

grand prix

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

 testator

-0.41

izio

-0.40

 alibi

-0.39

Mate

-0.39

PathVariable

-0.39

 actresses

-0.39

:{

-0.39

PPI

-0.38

 fates

-0.38

 cervix

-0.38

POSITIVE LOGITS

 Grand

1.15

Grand

1.09

 GRAND

0.98

GRAND

0.97

 grand

0.94

grand

0.93

 Grands

0.77

 Grande

0.76

Grande

0.71

グランド

0.68

Activations Density 0.004%

the term "grand" and its variations, indicating a focus on grandeur or significance

grand prix

No Comments

No Known Activations