INDEX

Explanations

instances of the word "display" and various forms of the word "over."

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

display item or rack

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

 CanadaChoose

-0.57

 Мексичка

-0.40

RegressionTest

-0.39

 ویکی‌پدی

-0.38

<eos>

-0.38

 Photocase

-0.37

);

-0.36

 незавершена

-0.36

TagMode

-0.34

POSITIVE LOGITS

 display

1.19

 Display

0.80

 exhibit

0.79

 displayed

0.79

BibitemShut

0.76

 DISPLAY

0.76

 displays

0.75

 displaying

0.69

bibnamefont

0.65

étoit

0.65

Activations Density 0.013%

instances of the word "display" and various forms of the word "over."

display item or rack

No Comments

No Known Activations