INDEX

Explanations

phrases and contexts related to opinions and decision-making

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

Preceding words to specific verbs

np_acts-logits-general · gemini-2.0-flash

academic and technical terms

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

 potentiel

-0.32

algo

-0.29

 conducteur

-0.28

 faldas

-0.28

 joueur

-0.28

Algo

-0.27

 algo

-0.27

yyr

-0.26

 réfé

-0.26

仗

-0.26

POSITIVE LOGITS



0.73

WebElementEntity

0.66

省市镇

0.64

脚注の使い方

0.60

-------------</

0.58

TypedDataSet

0.57

 الرياضيه

0.56

Fordítás

0.54

 IFTT

0.53

WebVitals

0.52

Activations Density 0.408%

phrases and contexts related to opinions and decision-making

Preceding words to specific verbs

academic and technical terms

No Comments

No Known Activations