INDEX

Explanations

numerical identifiers and mathematical notations commonly found in academic papers

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

mathematics subject classification codes

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

ſicht

-0.84

 Geiſt

-0.79

ロウィン

-0.78

 dieſen

-0.77

 deſſen

-0.77

ésultats

-0.77

 dieſem

-0.76

 ſeiner

-0.76

 témoig

-0.75

ſſung

-0.75

POSITIVE LOGITS

0.27

，

0.26

android

0.24

évaluateur

0.24

mesine

0.24

ViewImports

0.23

0.22

ében

0.22

Mad

0.21

band

0.21

Activations Density 0.004%

numerical identifiers and mathematical notations commonly found in academic papers

mathematics subject classification codes

No Comments

No Known Activations