INDEX

Explanations

elements related to legal sentences and judgments

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

consecutive versus concurrent sentences

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

Xna

-0.49

 surla

-0.48

formik

-0.47

存于互联网档案馆

-0.46

 فريبيس

-0.46

UnitTesting

-0.45

Cyfeiriadau

-0.44

 IERC

-0.43

evre

-0.43

Ծանոթ

-0.43

POSITIVE LOGITS

RenderAtEndOf

0.55

yntaxException

0.43

centralwidget

0.42

 SwitchCompat

0.41

BagConstraints

0.39

belajaran

0.38

execSQL

0.38

 Wicidata

0.37

SupportActionBar

0.36

ParallelGroup

0.36

Activations Density 0.065%

elements related to legal sentences and judgments

consecutive versus concurrent sentences

No Comments

No Known Activations