INDEX

Explanations

references to siblings, particularly sisters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

AutoField

-0.70

DoubleQuotes

-0.62

はじめに

-0.61

BeginContext

-0.61

SharedDtor

-0.59

phylactic

-0.54

مقاله

-0.53

 resourceCulture

-0.52

 Wikispecies

-0.52

 familiar

-0.52

POSITIVE LOGITS

Sisters

0.86

 Sisters

0.84

 sisters

0.84

Sister

0.82

sister

0.81

sisters

0.80

 Sister

0.80

 sister

0.73

 Winaray

0.69

ISTER

0.69

Activations Density 0.059%

references to siblings, particularly sisters

No Comments

No Known Activations