INDEX

Explanations

elements related to physical descriptions and processes

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

linking nouns to verbs/qualifiers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ModelExpression

-0.69

AddTagHelper

-0.68

WriteTagHelper

-0.67

oneofs

-0.66

AndEndTag

-0.65

๫

-0.65

󠀠

-0.65

 تضيفلها

-0.63

BeginContext

-0.61

 فريبيس

-0.61

POSITIVE LOGITS

Fordítás

0.31

stateMutability

0.29

ManyToOne

0.24

最快更新

0.23

പ

0.23

和

0.22

 Series

0.22

 liqu

0.22

 кри

0.21

人民共和国

0.21

Activations Density 0.114%

elements related to physical descriptions and processes

linking nouns to verbs/qualifiers

No Comments

No Known Activations