INDEX

Explanations

phrases and terms related to legal processes and concepts

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

legal terms and verbs

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

featureID

-0.64

 disambiguazione

-0.63

EDEFAULT

-0.53

 kasarigan

-0.53

KURZBESCHREIBUNG

-0.51

httphttps

-0.50

 HasFactory

-0.50

 Signalez

-0.50

 ModelExpression

-0.50

IndentedString

-0.49

POSITIVE LOGITS

 COMERC

0.37

 Roskov

0.36

Panamoan

0.35

 Ruß

0.35

InstrumentedTest

0.35

0.32

скать

0.32

 Hügel

0.31

Nen

0.31

Otro

0.30

Activations Density 0.090%

phrases and terms related to legal processes and concepts

legal terms and verbs

No Comments

No Known Activations