INDEX

Explanations

phrases related to numerical comparisons and equivalences

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

equivalence and proportion

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

LookAnd

-0.59

tagHelperRunner

-0.56

 betweenstory

-0.54

-0.50

verifyException

-0.50

+#+

-0.47

eze

-0.45

 ProtoMessage

-0.43

verwijspagina

-0.42

FG

-0.40

POSITIVE LOGITS

brigens

0.54

 equivalent

0.52

equivalent

0.51

相当于

0.47

 doubling

0.47

acamata

0.47

Equivalent

0.46

 equivale

0.45

 whopping

0.45

 desperdic

0.45

Activations Density 0.029%

phrases related to numerical comparisons and equivalences

equivalence and proportion

No Comments

No Known Activations