INDEX

Explanations

references to specific online platforms or links

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

hacker news and free republic

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

EndGlobalSection

-0.66

+#+#

-0.65

Personensuche

-0.64

GEBURTSDATUM

-0.60

WriteTagHelper

-0.59

 snippetHide

-0.57

OGND

-0.57

:✨

-0.56

ագրություններ

-0.52

الإنجليزية

-0.52

POSITIVE LOGITS

 okuyayım

0.49

bkz

0.35

Erreferentziak

0.32

 mCurrent

0.31

 interesan

0.30

 screenWidth

0.30

ptid

0.29

uminio

0.29

 servidores

0.29

❋

0.29

Activations Density 0.024%

references to specific online platforms or links

hacker news and free republic

No Comments

No Known Activations