INDEX

Explanations

patterns related to function and expression definitions in programming

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

code syntax punctuation

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

zzleHttp

-0.44

цездатний

-0.42

 للمعارف

-0.42

хьтан

-0.41

 PeEnEo

-0.39

 cherchés

-0.36

Autoritní

-0.36

esez

-0.35

ंदीखरीदारी

-0.34

 kaarangay

-0.34

POSITIVE LOGITS

 يتيمه

0.51

principalTable

0.50

richTextPanel

0.43

verwijspagina

0.43

مراجع

0.42

 designs

0.42

 mixes

0.42

 campione

0.42

 Versions

0.41

kuit

0.41

Activations Density 0.052%

patterns related to function and expression definitions in programming

code syntax punctuation

No Comments

No Known Activations