INDEX

Explanations

technical terms and actions related to software functionality and troubleshooting

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

application launch or update

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

elemField

-0.55

⅘

-0.48

RTEX

-0.47

parsedMessage

-0.46

/*

-0.44

FindBy

-0.44

HostException

-0.43

makeConstraints

-0.43

ParseTree

-0.42

/***/

-0.42

POSITIVE LOGITS

MLLoader

0.46

UserScript

0.46

FormTagHelper

0.45

 utafitiHapana

0.44

 contigo

0.41

 ligiloj

0.40

 springfox

0.38

 jspb

0.38

 opération

0.37

 operação

0.37

Activations Density 0.034%

technical terms and actions related to software functionality and troubleshooting

application launch or update

No Comments

No Known Activations