INDEX

Explanations

references to Quebec and its related regions or institutions

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

Québec, Montreal, French regions

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

 Colour

-0.53

 Colours

-0.50

 Behaviour

-0.49

tagHelperRunner

-0.48

Colour

-0.47

 AssemblyTitle

-0.47

Colours

-0.47

 للمعارف

-0.47

英国

-0.47

Autoritní

-0.47

POSITIVE LOGITS

 Quebec

0.50

 Haiti

0.50

 Haitian

0.50

Quebec

0.48

Haiti

0.46

 Montreal

0.46

Montreal

0.45

 Canadiens

0.42

Québec

0.40

 Louisiana

0.40

Activations Density 0.015%

references to Quebec and its related regions or institutions

Québec, Montreal, French regions

No Comments

No Known Activations