INDEX

Explanations

specific tokens and formatting in programming or code snippets

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

Code comments or headers

np_acts-logits-general · gemini-2.0-flash

// LLVM Printing Pass

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

BibitemShut

-0.51

angliski

-0.48

'{@

-0.47

 himſelf

-0.46

 Bienen

-0.44

abetes

-0.44

Obrázky

-0.43

@"/

-0.43

 ویکی‌آمباردا

-0.43

 bezeichneter

-0.43

POSITIVE LOGITS

colgroup

0.51

Роди

0.50

GEBURTS

0.50

ISupport

0.47

Wob

0.41

 wobble

0.40

 McKinley

0.39

MemoryWarning

0.39

 Ports

0.39

Poe

0.38

Activations Density 0.054%

specific tokens and formatting in programming or code snippets

Code comments or headers

// LLVM Printing Pass

No Comments

No Known Activations