INDEX

Explanations

web addresses or URLs

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

URLs and questions

np_acts-logits-general · gemini-2.0-flash

www. domains

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IntoConstraints

-0.91

rrggbb

-0.88

 Roskov

-0.86

 disambiguazione

-0.84

verwijspagina

-0.84

OGND

-0.82

 surla

-0.82

 autorytatywna

-0.81

 Lightboxes

-0.80

Aholisi

-0.80

POSITIVE LOGITS

 vermelhas

0.37

ToFit

0.36

 giác

0.34

kra

0.34

ña

0.32

án

0.32

laş

0.32

getElementById

0.31

 căng

0.30

étend

0.30

Activations Density 0.035%

web addresses or URLs

URLs and questions

www. domains

No Comments

No Known Activations