INDEX

Explanations

phrases related to features and benefits of products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 coucher

-0.65

 inalámb

-0.61

 esimer

-0.54

 splet

-0.53

 olsun

-0.52

 língua

-0.50

 meille

-0.50

 dieux

-0.50

pozorn

-0.49

 salvación

-0.49

POSITIVE LOGITS

 مشين

0.78

parsedMessage

0.70

MessageTagHelper

0.66

таратура

0.66

AsUp

0.62

Джерела

0.61

oredCriteria

0.60

awaiter

0.59

XmlType

0.59

tagHelperRunner

0.55

Activations Density 0.026%

phrases related to features and benefits of products

No Comments

No Known Activations