INDEX

Explanations

prompts for user input in code or instruction contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

IUrlHelper

-0.71

endphp

-0.61

makeText

-0.59

UserScript

-0.58

SOUNDBITE

-0.54

 nonUne

-0.54

onActivityResult

-0.52

 PackageManager

-0.51

QMetaType

-0.50

WebElementEntity

-0.48

POSITIVE LOGITS

tains

0.56

 Entry

0.54

taining

0.54

 into

0.51

tain

0.51

 Entries

0.49

way

0.48

Ent

0.47

tainer

0.47

PRISE

0.47

Activations Density 0.116%

prompts for user input in code or instruction contexts

No Comments

No Known Activations