INDEX

Explanations

terms related to processes, functionality, and application in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EDEFAULT

-0.72

 AssemblyCompany

-0.71

 ExecuteAsync

-0.65

Jeografia

-0.64

CloseOperation

-0.63

RegistryLite

-0.61

parsedMessage

-0.60

ThroughAttribute

-0.60

хьтан

-0.60

脚注の使い方

-0.60

POSITIVE LOGITS

tool

0.33

 средство

0.33

 strategi

0.33

 tool

0.32

validate

0.31

animate

0.31

finish

0.30

elesai

0.30

 some

0.30

 Mỹ

0.30

Activations Density 0.487%

terms related to processes, functionality, and application in various contexts

No Comments

No Known Activations