INDEX

Explanations

references to scripts or code alterations in a technical context

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

ChildScrollView

-0.50

AndEndTag

-0.48

dataclass

-0.43

synthesize

-0.41

Matchers

-0.40

 EconPapers

-0.40

homonymie

-0.39

uny

-0.39

 fulfilling

-0.38

 استنادى

-0.38

POSITIVE LOGITS

 purpoſe

0.55

 betweenstory

0.50

ſelves

0.46

transQ

0.46

 myſelf

0.46

 houſe

0.45

cesis

0.44

AutoModerator

0.43

 Verſ

0.43

 Reſ

0.42

Activations Density 0.004%

references to scripts or code alterations in a technical context

No Comments

No Known Activations