INDEX

Explanations

numerical representations indicating rank or significance associated with countries or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

<eos>

-0.52

?).

-0.51

挖

-0.50

 tuong

-0.50

 impea

-0.49

Had

-0.48

 Göttingen

-0.48

 occidentalis

-0.48

 capri

-0.48

 Mainz

-0.47

POSITIVE LOGITS

 числе

0.86

IsContent

0.80

WebServlet

0.80

AddTagHelper

0.77

CppMethod

0.75

setVerticalGroup

0.74

+#+#

0.73

@"/

0.71

ształ

0.71

 among

0.70

Activations Density 0.200%

numerical representations indicating rank or significance associated with countries or categories

No Comments

No Known Activations