INDEX

Explanations

references to popularity and trends

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

esis

-0.56

Autoritní

-0.49

 bandages

-0.49

UnitTesting

-0.48

 gneiss

-0.48

 cladding

-0.46

äch

-0.46

 endometri

-0.46

rophoresis

-0.45

 Fence

-0.45

POSITIVE LOGITS

 POPULAR

1.16

Popular

1.12

 Popular

1.12

 popular

1.11

popular

1.09

 populares

0.94

 popularidad

0.90

 popularity

0.89

 populaire

0.87

 Popularity

0.86

Activations Density 0.056%

references to popularity and trends

No Comments

No Known Activations