INDEX

Explanations

references to athletic competitions and achievements

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

Category names

np_acts-logits-general · gemini-2.0-flash

Category:

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

 AspNetCore

-0.54

Становништво

-0.51

 AssemblyCulture

-0.51

 ProtoMessage

-0.50

 للاسماء

-0.47

Izvori

-0.47

 Italij

-0.46

 newOwner

-0.45

Gön

-0.45

 escalating

-0.44

POSITIVE LOGITS

Diweddarwch

0.43

 bursa

0.30

his

0.30

 Administrativna

0.30

 Numerade

0.29

nub

0.29

社

0.28

 impar

0.28

RouterModule

0.28

EClass

0.28

Activations Density 0.802%

references to athletic competitions and achievements

Category names

Category:

No Comments

No Known Activations