INDEX

Explanations

references to lengthy or long arguments and comments

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

long answers or texts

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

AddTagHelper

-0.57

 EconPapers

-0.57

IUrlHelper

-0.56

 isKindOfClass

-0.50

ništvo

-0.48

IRQn

-0.48

 NSCoder

-0.48

openConnection

-0.47

abstractmethod

-0.47

Ecotoxicity

-0.47

POSITIVE LOGITS

 lengthy

0.55

 panjang

0.47

ยาว

0.42

writeByte

0.42

 Longest

0.42

 detailed

0.40

ChildScrollView

0.39

 compleja

0.38

 longest

0.37

 uzun

0.37

Activations Density 0.026%

references to lengthy or long arguments and comments

long answers or texts

No Comments

No Known Activations