INDEX

Explanations

references to NPR and its programming

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ButtonStyle

-0.53

 adver

-0.47

 Generales

-0.46

شار

-0.46

 Paper

-0.46

 Star

-0.46

raire

-0.45

 publish

-0.45

 Rite

-0.44

 Allociné

-0.44

POSITIVE LOGITS

NPR

0.94

NPR

0.78

PBS

0.76

Personendaten

0.73

addContainerGap

0.69

NHK

0.69

PBS

0.69

 חיצוניים

0.66

 BoxDecoration

0.65

 MaterialApp

0.65

Activations Density 0.180%

references to NPR and its programming

No Comments

No Known Activations