INDEX

Explanations

specific proper nouns and scientific terminologies

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

before the "orium" suffix

np_acts-logits-general · gemini-2.0-flash

prefixes of named entities or categories

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Carrière

-0.31

ˌ

-0.29

first

-0.29

 kiin

-0.28

 kysy

-0.28

 nyky

-0.28

capai

-0.28

context

-0.28

<h2>

-0.28

 pitä

-0.27

POSITIVE LOGITS

Datuak

0.86

ViewImports

0.83

 Paglinawan

0.79

Hentet

0.73

 препратки

0.73

 nakalista

0.71

AxisAlignment

0.70

 وتسجيلات

0.69

 विश्वसनीयता

0.68

 سكانية

0.68

Activations Density 1.133%

specific proper nouns and scientific terminologies

before the "orium" suffix

prefixes of named entities or categories

No Comments

No Known Activations