INDEX

Explanations

quantitative comparisons and relationships

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ConstraintMaker

-0.52

 disambiguazione

-0.49

 avoient

-0.47

Hauptartikel

-0.46

 Millet

-0.44

Schicken

-0.43

WireFormat

-0.43

estacks

-0.43

mobil

-0.41

 snippetHide

-0.41

POSITIVE LOGITS

GraphicsUnit

0.59

出版年

0.54

AndEndTag

0.45

Geplaatst

0.44

tvguidetime

0.39

 ModelExpression

0.39

 CreateTagHelper

0.36

########.

0.35

 صوتيه

0.35

 linkovi

0.35

Activations Density 0.218%

quantitative comparisons and relationships

No Comments

No Known Activations