INDEX

Explanations

occurrences of user-related references and programming structures

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

Tokens preceding colons or brackets

np_acts-logits-general · gemini-2.0-flash

which followed by a description

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

XmlAccessorType

-0.40

ngiliz

-0.35

DockStyle

-0.32

-0.31

RefNanny

-0.29

-0.27

 blessing

-0.27

 tibi

-0.27

 metav

-0.26

POSITIVE LOGITS

Tikang

1.10

 kasarigan

1.04

httphttps

0.93

Rüyada

0.92



0.88

出版年

0.88

 autorytatywna

0.87

expandindo

0.87

rungsseite

0.86

 Италијани

0.86

Activations Density 1.273%

occurrences of user-related references and programming structures

Tokens preceding colons or brackets

which followed by a description

No Comments

No Known Activations