INDEX

Explanations

references to specific types or labels

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

questions or lists

np_acts-logits-general · gemini-2.0-flash

common code elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 érde

-0.43

 verein

-0.40

 speciali

-0.40

 populaires

-0.39

 Folge

-0.39

wax

-0.39

אב

-0.39

 populaire

-0.39

 Sociedade

-0.38

 doux

-0.38

POSITIVE LOGITS

 esternos

1.22

 Мексичка

0.92

contentLoaded

0.92

Geplaatst

0.87

webElementGuid

0.84

IUrlHelper

0.84

MENAFN

0.82

SaveChangesAsync

0.82

tagHelperRunner

0.81

 saites

0.80

Activations Density 0.222%

references to specific types or labels

questions or lists

common code elements

No Comments

No Known Activations