INDEX

Explanations

program structure and specific command usage in coding

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

 مرئيه

-0.71

IUrlHelper

-0.59

weightedMode

-0.59

↹

-0.59

ChildIndex

-0.59

évaluateur

-0.57

acuzzi

-0.53

ViewFeatures

-0.52

elemField

-0.52

AddTagHelper

-0.52

POSITIVE LOGITS

basicConfig

0.37

 amarilla

0.36

 Escolar

0.33

 esimer

0.33

 benar

0.33

 combined

0.33

zeba

0.33

zerű

0.32

 Jüng

0.32

 Castilla

0.32

Activations Density 0.385%

program structure and specific command usage in coding

No Comments

No Known Activations