INDEX

Explanations

sections of code that relate to error handling or code comments

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

code related to CWE

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

ftagPool

-0.34

 wajah

-0.30

sig

-0.30

Tuple

-0.30

質

-0.29

dataclass

-0.28

styleType

-0.27

፩

-0.27

PopupMenu

-0.27

 dandy

-0.27

POSITIVE LOGITS

httphttps

0.74

msgTypes

0.72

VersionUID

0.71

 EClass

0.66

DockStyle

0.64

EndContext

0.63

 nahilalakip

0.62

ContentLoaded

0.61

principalTable

0.61

MigrationBuilder

0.59

Activations Density 0.001%

sections of code that relate to error handling or code comments

code related to CWE

No Comments

No Known Activations