INDEX

Explanations

variations and expressions of degree or intensity

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

bit followed by descriptor

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

 dennoch

-0.44

美しく

-0.42

 demeure

-0.41

 Ibid

-0.40

 upon

-0.40

 Xamarin

-0.40

 myriad

-0.40

多くの

-0.39

zaam

-0.38

依旧

-0.38

POSITIVE LOGITS



0.77

Anybody

0.71

everybody

0.69

setVerticalGroup

0.67

 Anybody

0.66

 anybody

0.65

RectangleBorder

0.64

 bloke

0.63

 aeroplane

0.63

 stuff

0.63

Activations Density 0.022%

variations and expressions of degree or intensity

bit followed by descriptor

No Comments

No Known Activations