INDEX

Explanations

references to news agencies or specific media outlets

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

AFP

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

 betweenstory

-0.61

fromnode

-0.51

AddTagHelper

-0.51

onViewCreated

-0.50

 مشين

-0.49

httphttps

-0.49

AutoScaleMode

-0.45

 defaultstate

-0.43

 архивлан

-0.42

 FetchType

-0.42

POSITIVE LOGITS

ⓧ

0.62

 disambiguazione

0.41

gameserver

0.41

 Impf

0.40

AFP

0.40

AFP

0.40

 eenige

0.40

SpringBootTest

0.40

RegistryLite

0.39

sspiel

0.39

Activations Density 0.004%

references to news agencies or specific media outlets

AFP

No Comments

No Known Activations