INDEX

Explanations

definitions and explanations of concepts or terms

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

defining or explaining what something is

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

 BrowserModule

-0.66

 TestBed

-0.66

 ویکی‌پدی

-0.65

تقاوى

-0.65

 Meksiku

-0.64

optionalTypeArgs

-0.63

 queſta

-0.63

/**

-0.63

verwijspagina

-0.61

最快更新

-0.61

POSITIVE LOGITS

er

0.39

 event

0.39

is

0.38

 bild

0.36

 object

0.36

 idea

0.35

 king

0.35

air

0.35

 exactly

0.35

 Notice

0.34

Activations Density 0.043%

definitions and explanations of concepts or terms

defining or explaining what something is

No Comments

No Known Activations