INDEX

Explanations

technical terms related to compatibility and installation issues

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

Tokens after punctuation or specific words

np_acts-logits-general · gemini-2.0-flash

only HP

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

 focused

-0.42

 sensitization

-0.35

 social

-0.34

 émissions

-0.32

 focusing

-0.32

 focus

-0.32

WithData

-0.32

 emot

-0.32

 dilution

-0.32

focused

-0.32

POSITIVE LOGITS

fjspx

0.78

 AssemblyCulture

0.69

DockStyle

0.69

 ujednoznacz

0.69

bootstrapcdn

0.64

ftagPool

0.64

ModelForm

0.62

 ProtoMessage

0.62

 CanadaChoose

0.60

الإنجليزية

0.59

Activations Density 0.190%

technical terms related to compatibility and installation issues

Tokens after punctuation or specific words

only HP

No Comments

No Known Activations