INDEX

Explanations

references to advice and appreciation in comments

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

your post

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

addPreferredGap

-0.37

 assets

-0.34

 cárcel

-0.33

 Handlungen

-0.32

 investigação

-0.32

ิ้ง

-0.30

 ägg

-0.30

綢

-0.30

 Vermögens

-0.30

 vlasy

-0.30

POSITIVE LOGITS

IndentedString

0.80

 estekak

0.69

 initComponents

0.65

 BoxFit

0.63

.*")]

0.63

 referrerpolicy

0.63

 ब्रेकडाउन

0.62

MessageTagHelper

0.61

UnusedPrivate

0.61

setupUi

0.60

Activations Density 0.102%

references to advice and appreciation in comments

your post

No Comments

No Known Activations