INDEX

Explanations

intensifiers or modifiers that express varying degrees of quality or quantity

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

most/more/very/less followed by adjective

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

nothing

-0.25

GEBURTSDATUM

-0.25

 falsa

-0.25

 there

-0.24

 prêtres

-0.24

 montagnes

-0.23

exitRule

-0.23

 چیزی

-0.23

foglal

-0.23

 gravidez

-0.22

POSITIVE LOGITS

 unſ

0.74

 Verſ

0.73

 myſelf

0.72

 ་་

0.71

 highly

0.71

niſſe

0.71

 ſind

0.71

ſelves

0.69

脚注の使い方

0.69

 Athenians

0.69

Activations Density 0.037%

intensifiers or modifiers that express varying degrees of quality or quantity

most/more/very/less followed by adjective

No Comments

No Known Activations