INDEX

Explanations

numerous quantifiable attributes and components within a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

Hochspringen

-0.42

="@+

-0.38

nestjs

-0.33

 subsequent

-0.33

bcryptjs

-0.31

ộp

-0.30

PageRoute

-0.30

 Subsequent

-0.29

astify

-0.29

 anum

-0.28

POSITIVE LOGITS

 finally

4.56

finally

4.00

 Finally

3.95

Finally

3.91

 finalmente

3.72

 enfin

3.59

 наконец

3.58

 FINALLY

3.41

 Finalmente

3.30

Finalmente

3.23

Activations Density 1.133%

numerous quantifiable attributes and components within a context

No Comments

No Known Activations