INDEX

Explanations

phrases indicating mathematical assumptions or conditions regarding positivity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

FAILED

-0.44

chin

-0.43

ed

-0.42

 trebui

-0.41

XmlAccessorType

-0.40

et

-0.38

 Korn

-0.38

 kém

-0.38

Shaft

-0.38

lys

-0.38

POSITIVE LOGITS

 propOrder

0.72

.*")]

0.69

 محفوظة

0.68

 nakalista

0.67

+:+

0.67

 resourceCulture

0.65

ISupport

0.65

nodoc

0.64

usermodel

0.63

 comigo

0.62

Activations Density 0.009%

phrases indicating mathematical assumptions or conditions regarding positivity

No Comments

No Known Activations