INDEX

Explanations

details about electronic devices, particularly focusing on specifications and models

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

product names and models

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

 vectorielles

-0.43

 Christian

-0.37

 fær

-0.36

 science

-0.35

assertThat

-0.34

 cristian

-0.34

fæ

-0.33

 Brasilien

-0.32

 commerciales

-0.32

 culture

-0.32

POSITIVE LOGITS

IntoConstraints

0.68

MessageTagHelper

0.67

ValueStyle

0.66

FontOfSize

0.64

bootstrapcdn

0.61

pidou

0.60

AsUp

0.60

OGND

0.59

ConstraintMaker

0.59

URLException

0.59

Activations Density 0.012%

details about electronic devices, particularly focusing on specifications and models

product names and models

No Comments

No Known Activations