INDEX

Explanations

articles and nouns that denote identity or classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 transactions

-0.46

 condena

-0.40

 detalla

-0.40

 transcur

-0.40

 encont

-0.40

 requests

-0.40

 Metcalf

-0.40

 odpo

-0.39

 trips

-0.39

 femminile

-0.39

POSITIVE LOGITS

PerformLayout

0.68

帖最后由

0.63

 kasarigan

0.59

AutoresizingMask

0.51

 citizen

0.51

 Majefty

0.49

eratops

0.49

 myſelf

0.48

 Citizen

0.48

 AssemblyCompany

0.47

Activations Density 0.334%

articles and nouns that denote identity or classification

No Comments

No Known Activations