INDEX

Explanations

elements that indicate effectiveness and characteristics of testing and evaluation processes

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

combines or manipulates elements

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

:✨

-0.54

Rüyada

-0.48

ureusement

-0.47

omentos

-0.43

 Tatsache

-0.43

 المعيارى

-0.43

Specyfikacja

-0.42

enterOuterAlt

-0.42

 ويكيپيديا

-0.41

Ӕ

-0.41

POSITIVE LOGITS

RenderAtEndOf

0.70

TagMode

0.52

 ModelExpression

0.44

таратура

0.44

رشف

0.40

ſy

0.40

 ſur

0.40

 ujednoznacz

0.40

 loue

0.39

 indi

0.38

Activations Density 0.071%

elements that indicate effectiveness and characteristics of testing and evaluation processes

combines or manipulates elements

No Comments

No Known Activations