INDEX

Explanations

specific sequences of alphanumeric characters and formatting typically associated with technical or programming code

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

user input or interaction

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

-0.66

+#+#

-0.66

 Италијани

-0.57

AnchorTagHelper

-0.52

 propOrder

-0.51

:✨

-0.51

 "..\..\..\

-0.50

#+#

-0.50

XmlAccessType

-0.49

httphttps

-0.49

POSITIVE LOGITS

 getPlayer

0.40

bcryptjs

0.38

Obrázky

0.35

 Pulaski

0.34

 MARSH

0.33

 parle

0.32

replaceable

0.32

AppMethodBeat

0.32

parsedMessage

0.32

 estudante

0.32

Activations Density 0.001%

specific sequences of alphanumeric characters and formatting typically associated with technical or programming code

user input or interaction

No Comments

No Known Activations