INDEX

Explanations

references to scholarly work, citing authors and their contributions within academic contexts

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

academic citations and authors

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

AutoresizingMask

-0.52

ChildScrollView

-0.52

DoubleQuotes

-0.48

WebElementEntity

-0.47

IntoConstraints

-0.43

[]:

-0.40

IBarButtonItem

-0.40

</thead>

-0.39

 breech

-0.39

Cyfeiriadau

-0.39

POSITIVE LOGITS

 Studien

0.42

鶫

0.42

oracle

0.41

 публика

0.40

apropri

0.39

 Arbeiten

0.38

 Verfügung

0.38

脚注の使い方

0.38

 ชั้น

0.38

ściej

0.37

Activations Density 0.324%

references to scholarly work, citing authors and their contributions within academic contexts

academic citations and authors

No Comments

No Known Activations