INDEX

Explanations

conjunctions and phrases indicating relationships or connections between ideas

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

improving outcomes and reducing issues

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

kne

-0.39

 jäsen

-0.39

lug

-0.38

 ää

-0.36

 coroa

-0.36

zett

-0.35

 Brus

-0.35

 kokona

-0.35

 frigid

-0.35

 hook

-0.35

POSITIVE LOGITS

новниш

0.60

stylers

0.56

xase

0.56

 typelib

0.54

ValueStyle

0.54

 Autorizaciones

0.53

 poffible

0.53

Zeneca

0.53

hance

0.51

campal

0.50

Activations Density 0.069%

conjunctions and phrases indicating relationships or connections between ideas

improving outcomes and reducing issues

No Comments

No Known Activations