INDEX

Explanations

references to sequences and structures, particularly in narrative contexts

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

introduces explanation or description

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

 enfans

-0.62

 desmotivaciones

-0.56

 florales

-0.53

 morire

-0.53

 mármol

-0.52

oredCriteria

-0.50

 burbujas

-0.49

 démocr

-0.49

 Türkei

-0.49

 Sünde

-0.49

POSITIVE LOGITS

 nahilalakip

0.55

 oprot

0.46

Diwedd

0.37

inf

0.36

клопе

0.36

 pinn

0.34

 CURIAM

0.34

agu

0.34

پوش

0.33

Rujuakan

0.33

Activations Density 0.193%

references to sequences and structures, particularly in narrative contexts

introduces explanation or description

No Comments

No Known Activations