INDEX

Explanations

positive statements or affirmations

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

technical, scientific, or legal contexts

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

getClassLoader

-0.45

mtable

-0.34

 îmb

-0.33

 JADX

-0.33

Население

-0.32

ORES

-0.31

äť

-0.31

 Hickman

-0.31

 missione

-0.31

 désolés

-0.30

POSITIVE LOGITS

parsedMessage

0.59

BufferException

0.56

 utafitiHapana

0.55

expandindo

0.51

BagConstraints

0.50

writeField

0.50

IndentedString

0.48

sidemargin

0.47

styleType

0.47

 cherchés

0.47

Activations Density 0.309%

positive statements or affirmations

technical, scientific, or legal contexts

No Comments

No Known Activations