INDEX

Explanations

ordinal numbers and their rankings

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

numerical rankings and positions

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

ordinal followed by superlative

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

 betweenstory

-0.83

MigrationBuilder

-0.80

 transfieras

-0.78

جغرافيا

-0.77

GTCX

-0.73

IUrlHelper

-0.73

Personensuche

-0.73

 referenties

-0.72

encodeWith

-0.68

RenderAtEndOf

-0.67

POSITIVE LOGITS

 most

1.53

most

1.25

 best

1.25

 largest

1.22

 highest

1.17

 strongest

1.13

 longest

1.11

 biggest

1.08

best

1.07

 busiest

1.06

Activations Density 0.065%

ordinal numbers and their rankings

numerical rankings and positions

ordinal followed by superlative

No Comments

No Known Activations