INDEX

Explanations

references to research articles and their associated metadata

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Vege

-0.39

吞

-0.39

thme

-0.39

inout

-0.38

adhy

-0.38

荷

-0.37

umm

-0.37

vez

-0.36

amba

-0.36

脚注の使い方

-0.36

POSITIVE LOGITS

 PLOS

0.61

 PLoS

0.60

 defaultstate

0.58

évaluateur

0.52

OrNil

0.51

PreferredItem

0.51

titleMargin

0.50

出版年

0.48

 Biôgrafia

0.45

PLOS

0.45

Activations Density 1.202%

references to research articles and their associated metadata

No Comments

No Known Activations