INDEX

Explanations

quantitative data or numerical values in mathematical or scientific contexts

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

numbers and code constructs

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

LookAnd

-0.53

transQ

-0.48

postIndex

-0.45

 GoogleFonts

-0.44

fjspx

-0.44

UIControlState

-0.43

 <>",

-0.42

RegressionTest

-0.41

оргаш

-0.41

 stabilité

-0.40

POSITIVE LOGITS

TagMode

0.47

новништво

0.40

complexContent

0.39

Hud

0.39

typically

0.38

zion

0.37

__*/

0.35

Personendaten

0.35

Referències

0.35

EDEFAULT

0.34

Activations Density 0.045%

quantitative data or numerical values in mathematical or scientific contexts

numbers and code constructs

No Comments

No Known Activations