INDEX

Explanations

words and phrases that indicate high quality or high performance

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

 المعيارى

-0.64

 noires

-0.51

 Marav

-0.51

 celles

-0.49

 possibles

-0.48

väg

-0.48

 étoit

-0.47

dengar

-0.47

paper

-0.47

 enfans

-0.47

POSITIVE LOGITS

esModule

0.67

queryInterface

0.67

 quan

0.64

 grade

0.59

 renounced

0.57

libft

0.57

 color

0.56

legende

0.56

 Quan

0.56

ApiModelProperty

0.56

Activations Density 0.090%

words and phrases that indicate high quality or high performance

No Comments

No Known Activations