INDEX

Explanations

similarities in formatting or structure within the data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

تقاوى

-0.63

 Bouch

-0.49

$,

-0.45

IsInitialized

-0.44

 $_{\

-0.44

〞

-0.43

]?.

-0.43

$\$$

-0.43

styleType

-0.42

POSITIVE LOGITS

----------------

1.33

------

1.04

-----

1.04

--------------

1.02

---------

1.02

---

1.02

-----------

1.02

------------

1.01

----

1.01

--------

0.99

Activations Density 0.526%

similarities in formatting or structure within the data

No Comments

No Known Activations