INDEX

Explanations

sections and elements related to academic papers or presentations

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

section headings

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

########.

-0.56

 utafitiHapana

-0.48

 Administrativna

-0.48

 surla

-0.46

 Cloth

-0.46

WISH

-0.44

ɵɵ

-0.43

rachtet

-0.42

 razie

-0.41

Cloth

-0.41

POSITIVE LOGITS

setVerticalGroup

0.54

hoeddwyd

0.47

httphttps

0.42

ثيق

0.40

RuleContext

0.39

XmlAccessType

0.38

 StatefulWidget

0.36

 NSCoder

0.36

ScopeManager

0.35

kuu

0.35

Activations Density 0.001%

sections and elements related to academic papers or presentations

section headings

No Comments

No Known Activations