INDEX

Explanations

technical terms and symbols related to programming or coding

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

user

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

 χρόν

-0.39

apeno

-0.38

 still

-0.38

 action

-0.37

 actually

-0.36

 became

-0.36

 rimasto

-0.35

不已

-0.35

కు

-0.34

exécution

-0.34

POSITIVE LOGITS

 Мексичка

0.61

IUrlHelper

0.57

posedge

0.56

 Numerade

0.51

Билгалдахарш

0.51

 kasarigan

0.50

usepackage

0.50

hyrchwyd

0.49

postsleuth

0.49

ंदीखरीदारी

0.49

Activations Density 0.000%

technical terms and symbols related to programming or coding

user

No Comments

No Known Activations