INDEX

Explanations

articles and quantifiers that precede nouns

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

a followed by room descriptions

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

 EconPapers

-0.67

 ویکی‌پدیای

-0.50

 disambiguazione

-0.49

toFixed

-0.49

dération

-0.49

complexType

-0.49

simpleType

-0.49

 HasFactory

-0.48

SpringBootTest

-0.47

Reprodução

-0.47

POSITIVE LOGITS

 bathtub

0.44

 sunroof

0.42

 esternos

0.42

 cantile

0.41

 scarf

0.41

 kitchen

0.39

 barbe

0.39

 swivel

0.39

 bathroom

0.38

 grill

0.38

Activations Density 0.001%

articles and quantifiers that precede nouns

a followed by room descriptions

No Comments

No Known Activations