INDEX

Explanations

references to matchups or comparisons in competitive contexts

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

vs, versus, and, or

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

nodoc

-0.45

 needs

-0.38

 igény

-0.37

WriteBarrier

-0.36

ží

-0.35

VersionUID

-0.35

MigrationBuilder

-0.34

TestingModule

-0.34

AnchorStyles

-0.34

ighed

-0.34

POSITIVE LOGITS

verwijspagina

0.52

HasIndex

0.46

utilisons

0.46

 snippetHide

0.45

RenderAtEndOf

0.44

новништво

0.44

ensement

0.43

ⓧ

0.42

 Jinping

0.41

 gynhyrchwyd

0.40

Activations Density 0.048%

references to matchups or comparisons in competitive contexts

vs, versus, and, or

No Comments

No Known Activations