INDEX

Explanations

programming or coding-related elements

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

Japanese acknowledgement

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

 estekak

-0.68

'{@

-0.59

BindView

-0.54

INSEE

-0.53

oneofs

-0.53

SharedCtor

-0.53



-0.52

 propTypes

-0.52

bitField

-0.52

onViewCreated

-0.52

POSITIVE LOGITS

ibilidade

0.44

sibilidade

0.39

 civilización

0.35

 Nacionales

0.34

ิลปะ

0.34

 transparent

0.34

RegistryLite

0.33

 traduce

0.32

萝卜

0.32

 quyền

0.31

Activations Density 0.001%

programming or coding-related elements

Japanese acknowledgement

No Comments

No Known Activations