INDEX

Explanations

concepts related to evaluation and critique of representation and performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

verwijspagina

-0.28

 director

-0.25

texttt

-0.24

 supp

-0.24

 esternos

-0.23

?”

-0.23

 presa

-0.23

lapsingToolbar

-0.23

フォル

-0.23

che

-0.22

POSITIVE LOGITS

prepareStatement

0.70

onOptions

0.69

ImageContext

0.66

󠁴

0.66

 дописавши

0.66

XMLSchema

0.64

EDEFAULT

0.64

⬯

0.63

 tartalo

0.63

<unused42>

0.63

Activations Density 0.779%

concepts related to evaluation and critique of representation and performance

No Comments

No Known Activations