INDEX

Explanations

elements related to programming or technical instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

 they

-0.59

 itself

-0.47

 Itself

-0.46

umesc

-0.45

Hozzáférés

-0.43

They

-0.43

strze

-0.41

 himself

-0.41

 yours

-0.41

drawiam

-0.41

POSITIVE LOGITS

 ourselves

1.61

we

1.43

我们

1.21

我們

1.19

 мы

1.16

我们在

1.16

私たちは

1.14

 我们

1.09

 możemy

1.06

我们就

1.05

Activations Density 0.040%

elements related to programming or technical instructions

No Comments

No Known Activations