INDEX

Explanations

references to performance metrics and cost considerations in technical contexts

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

simple Network Manager

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

 <<<<<<<<<<<<<<

-0.48

قایناقلار

-0.46

 disambiguazione

-0.45

MenuInflater

-0.45

 تكبرها

-0.44

windowFixed

-0.42

 '\\;'

-0.42

httphttps

-0.41

 plais

-0.41

Comté

-0.41

POSITIVE LOGITS

 betweenstory

0.58

msgTypes

0.52

featureID

0.47

CloseOperation

0.43

 pinulongan

0.41

displayquote

0.41

Rohy

0.41

FieldBuilder

0.38

小山

0.38

+#+#

0.37

Activations Density 0.113%

references to performance metrics and cost considerations in technical contexts

simple Network Manager

No Comments

No Known Activations