INDEX

Explanations

terms and phrases related to computing and programming constructs

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

compile-time or pre-defined sequences

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

binder

-0.46

Zed

-0.46

mug

-0.45

 nahilalakip

-0.45

scra

-0.44

bills

-0.44

醐

-0.44

inSlope

-0.43

aplic

-0.43

Javadoc

-0.42

POSITIVE LOGITS

pre

0.50

 predefined

0.46

 predetermined

0.45

 AssemblyCulture

0.44

 Brücken

0.42

__(/*!

0.40

 antemano

0.38

zysty

0.37

 setuptools

0.37

 BrowserRouter

0.36

Activations Density 0.103%

terms and phrases related to computing and programming constructs

compile-time or pre-defined sequences

No Comments

No Known Activations