INDEX

Explanations

terms related to "MA" or specific references to marketing and product names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

AddTagHelper

-0.64

parsedMessage

-0.63

webElementXpaths

-0.58

tagHelperRunner

-0.55

 Aussicht

-0.51

verwijspagina

-0.48

 BorderSide

-0.48

:✨

-0.48

 nahilalakip

-0.48

IndentedString

-0.48

POSITIVE LOGITS

Ma

0.66

Ma

0.60

MA

0.59

 sure

0.55

intosh

0.49

ma

0.49

 mẽ

0.47

 Sure

0.46

Mav

0.44

MA

0.44

Activations Density 0.442%

terms related to "MA" or specific references to marketing and product names

No Comments

No Known Activations