INDEX

Explanations

sections of high numerical significance or emphasis

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

percentages and numerical data

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 AssemblyCulture

-0.83

 disambiguazione

-0.75

uxxxx

-0.73

NOPQRST

-0.70

+#+#

-0.69

 المعيارى

-0.67

featureID

-0.66

rungsseite

-0.66

########.

-0.65

 ویکی‌پدی

-0.65

POSITIVE LOGITS

 EnglishChoose

0.47

ajoz

0.34

xrLabel

0.32

 foran

0.31

见

0.31

 Arxivat

0.29

 uved

0.29

 pronon

0.28

 provi

0.28

Herkunft

0.28

Activations Density 0.429%

sections of high numerical significance or emphasis

percentages and numerical data

No Comments

No Known Activations