INDEX

Explanations

terms related to physical spaces or areas

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

space and tissue

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

 useDispatch

-0.54

 MainAxisSize

-0.48

 canadienne

-0.46

istoitu

-0.46

 numerically

-0.45

 getLogin

-0.45

 intellectuelle

-0.45

 inoxydable

-0.44

════════

-0.44

 sendok

-0.44

POSITIVE LOGITS

 surla

0.59

 park

0.47

 Fläche

0.46

 déchets

0.44

 espaces

0.43

 forêts

0.41

 greenery

0.41

 équipements

0.41

parian

0.41

 obce

0.40

Activations Density 0.007%

terms related to physical spaces or areas

space and tissue

No Comments

No Known Activations