INDEX

Explanations

sections and formatting elements often used in academic articles or papers

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

citations and figures links

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Personendaten

-0.85

 disambiguazione

-0.77

 tartalomajánló

-0.76

AndEndTag

-0.74

expandindo

-0.73

 autorytatywna

-0.72

SharedCtor

-0.67

TagMode

-0.65

يميديا

-0.63

 Савезне

-0.61

POSITIVE LOGITS

Zwe

0.35

obstacles

0.32

 Cloak

0.31

vil

0.30

OREM

0.30

 població

0.30

'{@

0.29

 frente

0.29

PLN

0.29

 auquel

0.29

Activations Density 0.050%

sections and formatting elements often used in academic articles or papers

citations and figures links

No Known Activations