INDEX

Explanations

LaTeX commands or syntax in the document

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

start of user turn

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

 يتيمه

-0.81

 صوتيه

-0.63

EndContext

-0.55

 Administrativna

-0.54

:✨

-0.54

 Meksiku

-0.52

 Italijani

-0.52

ckså

-0.51

 Wikimedijinoj

-0.51

 Infórmanos

-0.50

POSITIVE LOGITS

 giriş

0.47

An

0.46

hive

0.45

 document

0.45

">//

0.44

PRINT

0.44

kuri

0.43

An

0.43

ada

0.43

The

0.42

Activations Density 0.000%

LaTeX commands or syntax in the document

start of user turn

No Comments

No Known Activations