INDEX

Explanations

mentions of characters and their attributes in a narrative context

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

Kyutama

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

 doInBackground

-0.67

SBATCH

-0.56

 Skyrim

-0.52

featureID

-0.52

Mule

-0.52

 وتسجيلات

-0.52

TagMode

-0.51

KURZBESCHREIBUNG

-0.50

 mules

-0.50

ritsar

-0.49

POSITIVE LOGITS

キュア

0.39

 Vort

0.34

劝

0.34

 minute

0.33

 zost

0.33

 perif

0.33

 Koning

0.32

 perfon

0.32

ActionListener

0.32

 Cliquez

0.32

Activations Density 0.248%

mentions of characters and their attributes in a narrative context

Kyutama

No Comments

No Known Activations