INDEX

Explanations

specific structural components and details in descriptions

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

render HTML tags

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

 razonable

-0.36

 duele

-0.34

 arrêté

-0.34

 productivo

-0.33

 múltiple

-0.33

einander

-0.32

 marquées

-0.32

 moviliz

-0.32

 עוד

-0.31

 temprano

-0.31

POSITIVE LOGITS

MLLoader

0.73

ValueStyle

0.65

RenderAtEndOf

0.64

expandindo

0.63

AndEndTag

0.62

 للاسماء

0.61

 فريبيس

0.60

yntaxException

0.59

GOTREF

0.58

 الحره

0.57

Activations Density 0.105%

specific structural components and details in descriptions

render HTML tags

No Comments

No Known Activations