INDEX

Explanations

references to abstract concepts or ideas

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

loadModel

-0.87

omla

-0.84

FileNotFound

-0.84

endpush

-0.81

 GoogleFonts

-0.81

 Henson

-0.79

 بيها

-0.76

RunWith

-0.76

PMailer

-0.74

IBarButtonItem

-0.73

POSITIVE LOGITS

 abstracts

1.24

 Abstracts

1.08

 Abstract

1.00

 抽象

0.98

 abstract

0.97

Abstracts

0.92

Abstract

0.92

abstract

0.92

 abstracted

0.88

abstra

0.87

Activations Density 0.005%

references to abstract concepts or ideas

No Comments

No Known Activations