INDEX

Explanations

interrogative phrases or questions

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

French text, often in chat or help forums

np_acts-logits-general · gemini-2.0-flash

quelqu'un

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 betweenstory

-0.83

 CreateTagHelper

-0.79

 nahilalakip

-0.78

ArrowToggle

-0.73

windowFixed

-0.71

 ExecuteAsync

-0.70

setVerticalGroup

-0.68

 '\\;'

-0.65

PreferredItem

-0.64

 Walkover

-0.63

POSITIVE LOGITS

 aliqu

0.57

 Tujuan

0.54

存于互联网档案馆

0.49

lainen

0.49

onaceous

0.49

 RECORDING

0.48

 Kerja

0.48

 تانيه

0.48

 PhpStorm

0.47

 Calvo

0.47

Activations Density 0.084%

interrogative phrases or questions

French text, often in chat or help forums

quelqu'un

No Comments

No Known Activations