INDEX

Explanations

the start of sections, often labelled or categorized content

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

specific descriptive phrases

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

}{*}{}

-0.81

 trovar

-0.72

enumi

-0.69

GraphicsUnit

-0.65

SequentialGroup

-0.65

 newOwner

-0.65

 hemispheres

-0.64

 gekomen

-0.63

instan

-0.61



-0.61

POSITIVE LOGITS

0.61

low

0.52

 small

0.52

 high

0.50

 mixte

0.50

 kaarangay

0.48

 multi

0.47

 stanovnika

0.47

Ārējās

0.47

Activations Density 0.561%

the start of sections, often labelled or categorized content

specific descriptive phrases

No Comments

No Known Activations