INDEX

Explanations

programming or software-related terms and class structures

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

math, asio, ngx code snippets

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

 rulers

-0.31

TextBoxColumn

-0.31

 ruler

-0.31

 PROVID

-0.29

 Cler

-0.29

alth

-0.29

actionMode

-0.28

 Dickens

-0.28

avon

-0.28

перь

-0.28

POSITIVE LOGITS

 surla

1.13

httphttps

1.00

 intptr

0.97

tagHelperRunner

0.94

ValueStyle

0.80

 ویکی‌پدی

0.79

NameInMap

0.78

MigrationBuilder

0.76

 disponibilités

0.70

hyperref

0.69

Activations Density 0.062%

programming or software-related terms and class structures

math, asio, ngx code snippets

No Comments

No Known Activations