INDEX

Explanations

references to officials and authority figures in various contexts

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

preceding government/research roles

np_acts-logits-general · gemini-2.0-flash

responsible officials and researchers

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-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

RectangleBorder

-0.83

 Мексичка

-0.82

PyExc

-0.74

principalColumn

-0.71

webElementXpaths

-0.71

ISupport

-0.71

 समीक्षक

-0.70

 AppComponent

-0.69

RTLD

-0.69

 становника

-0.67

POSITIVE LOGITS

 responsible

0.81

 overseeing

0.76

 responsáveis

0.76

 responsables

0.72

 zufolge

0.71

 tarafından

0.70

 compét

0.70

 responsável

0.69

 chargés

0.66

 verantwoorde

0.65

Activations Density 0.492%

references to officials and authority figures in various contexts

preceding government/research roles

responsible officials and researchers

No Comments

No Known Activations