INDEX

Explanations

references to specific entities or actions in a conversation

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

May which why

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

новништво

-0.63

AndEndTag

-0.62

TagMode

-0.58

Tembelea

-0.57

 '\\;'

-0.56

 pinulongan

-0.51

featureID

-0.48

 CURIAM

-0.44

 Autorizaciones

-0.43

providedIn

-0.42

POSITIVE LOGITS

BufferException

0.58

0.40

 certeza

0.40

Бележки

0.40

CPtr

0.39

utsches

0.39

assertNotNull

0.38

 resourceCulture

0.38

記得

0.37

iguar

0.37

Activations Density 0.028%

references to specific entities or actions in a conversation

May which why

No Comments

No Known Activations