INDEX

Explanations

references to empirical experiments and their results

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 informée

-0.56

ſelf

-0.51

 autorytatywna

-0.49

RTEX

-0.46

 مشين

-0.46

BindView

-0.45

 houſe

-0.45

 ſtate

-0.45

TypedDataSet

-0.45

orianCalendar

-0.45

POSITIVE LOGITS

 toekomst

0.47

 results

0.46

 hasilnya

0.46

 experimental

0.46

 Ergebnisse

0.45

 wijze

0.45

 experimentally

0.44

MessageTagHelper

0.41

 TestBed

0.41

 superiority

0.41

Activations Density 1.305%

references to empirical experiments and their results

No Comments

No Known Activations