INDEX

Explanations

characteristics or elements that indicate identification or categorization within a structured format

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

numbers and code

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

 EconPapers

-0.55

 spéciaux

-0.50

ResponseWriter

-0.50

SharedCtor

-0.49

providedIn

-0.48

 AssemblyTitle

-0.47

 familiale

-0.46

 Autorisations

-0.46

RectangleBorder

-0.44

createCanvas

-0.44

POSITIVE LOGITS

 виправивши

0.48

 ReactDOM

0.42

Hentet

0.42

 noqa

0.41

LabelTagHelper

0.36

 distanciation

0.35

prefixer

0.35

 وتسجيلات

0.34

cestry

0.34

Abitanti

0.34

Activations Density 0.001%

characteristics or elements that indicate identification or categorization within a structured format

numbers and code

No Comments

No Known Activations