INDEX

Explanations

phrases associated with functionality and utility in technology-related discussions

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

useful and functional

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

相关文章

-0.39

umplimiento

-0.34

 related

-0.33

ヒント

-0.33

 ActionResult

-0.32

bir

-0.32

fed

-0.32

RTDA

-0.32

Werdegang

-0.32

 MenuView

-0.32

POSITIVE LOGITS

 berguna

0.60

 useful

0.56

 Useful

0.56

useful

0.55

 flexibility

0.52

 usefulness

0.52

Autoritní

0.52

 Савезне

0.51

Useful

0.51

 útiles

0.50

Activations Density 0.158%

phrases associated with functionality and utility in technology-related discussions

useful and functional

No Comments

No Known Activations