INDEX

Explanations

specific programming terms and structures, likely related to coding and software development

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

information related to Docker configurations or network settings

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

numbers, code, or programming constructs

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

 croix

-0.40

kri

-0.38

 adult

-0.37

 مع

-0.36

ind

-0.36

 comidas

-0.35

cru

-0.34

 sell

-0.34

en

-0.34

 proximité

-0.34

POSITIVE LOGITS

 Roskov

0.88

 Мексичка

0.87

IntoConstraints

0.85

PhysRev

0.81

 subnet

0.78

mask

0.76

 Jefus

0.76

جغرافيا

0.76

 AssemblyCulture

0.74

 mask

0.74

Activations Density 0.111%

specific programming terms and structures, likely related to coding and software development

information related to Docker configurations or network settings

numbers, code, or programming constructs

No Comments

No Known Activations