INDEX

Explanations

references to time periods, particularly the term "year"

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

are

-0.49

╚

-0.47

Děkuji

-0.47

for

-0.45

des

-0.45

day

-0.45

 finally

-0.44

or

-0.43

ámb

-0.43

')[

-0.43

POSITIVE LOGITS

 صوتيه

1.26

########.

1.08

DoubleQuotes

1.02

 الحره

1.00

BeginContext

0.99

tagHelperRunner

0.98

InjectAttribute

0.95

ViewFeatures

0.95

NameInMap

0.91

 AssemblyProduct

0.88

Activations Density 0.064%

references to time periods, particularly the term "year"

No Comments

No Known Activations