INDEX

Explanations

terms related to data processing and reconstruction in imaging

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

acquired or displayed

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

pushFollow

-0.45

 CURIAM

-0.44

WireFormatLite

-0.42

 محفوظة

-0.42

 nonUne

-0.41

JspWriter

-0.40

KURZBESCHREIBUNG

-0.40

ArgsConstructor

-0.39

tranno

-0.39

RUnlock

-0.38

POSITIVE LOGITS

 cleaner

0.44

kle

0.44

createServer

0.43

 Separ

0.42

 CLEAN

0.41

 Cleaner

0.40

cab

0.40

thums

0.40

 separately

0.40

Separ

0.40

Activations Density 0.181%

terms related to data processing and reconstruction in imaging

acquired or displayed

No Comments

No Known Activations