INDEX

Explanations

patterns related to numerical values and their associated operations or annotations

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

Exponents (e.g., 1.00E-04)

np_acts-logits-general · gemini-2.0-flash

scientific notation numbers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

-0.49

less

-0.45

 potra

-0.44

ore

-0.43

чин

-0.41

-0.40

-0.39

 bañ

-0.38

dav

-0.37

POSITIVE LOGITS

IntoConstraints

1.08

ंदीखरीदारी

1.02

providedIn

0.99

 فريبيس

0.94

 MainAxisSize

0.92

UnusedPrivate

0.92

enderror

0.89

AndEndTag

0.88

ReusableCell

0.86

Референце

0.85

Activations Density 0.291%

patterns related to numerical values and their associated operations or annotations

Exponents (e.g., 1.00E-04)

scientific notation numbers

No Comments

No Known Activations