INDEX

Explanations

specific symbols and formatting elements in a coding or markup environment

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

user prompts

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

 GlobalKey

-0.55

 WebDriverWait

-0.51

 invokingState

-0.47

LocationManager

-0.47

 propOrder

-0.46

 Persson

-0.44

SaveChangesAsync

-0.44

AntiForgeryToken

-0.43

 configureStore

-0.43

UnusedPrivate

-0.43

POSITIVE LOGITS

 الحره

0.54

CppMethod

0.49

 cleanest

0.44

 kasarigan

0.44

 esternos

0.43

encils

0.43

ьажоргаш

0.42

Allo

0.42

 صوتيه

0.41

 tartalomajánló

0.41

Activations Density 0.000%

specific symbols and formatting elements in a coding or markup environment

user prompts

No Comments

No Known Activations