INDEX

Explanations

the word "be" in various contexts, indicating a focus on the verb's usage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

íodo

-0.47

 prime

-0.45

 Tienes

-0.45

並沒有

-0.44

 получают

-0.43

开了

-0.42

 habet

-0.42

are

-0.42

 Partagez

-0.42

 predicts

-0.42

POSITIVE LOGITS

ValueStyle

1.34

SharedDtor

1.17

 Biôgrafia

1.11

Vidite

1.10

ConstraintMaker

1.10

AsUp

1.10

ToBe

1.09

\{\\

1.08

ьаж

1.08

tvguidetime

1.07

Activations Density 0.097%

the word "be" in various contexts, indicating a focus on the verb's usage

No Comments

No Known Activations