INDEX

Explanations

references to places or geographical locations, particularly related to the term "Hills."

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

<?

-0.51

 femininas

-0.50

PYX

-0.48

__":

-0.48

 CURIAM

-0.47

 HasFactory

-0.47

opsida

-0.46

BibitemShut

-0.46

 autorytatywna

-0.46

 كومونز

-0.45

POSITIVE LOGITS

 Hills

0.72

 HILLS

0.64

stamp

0.61

 stamp

0.61

 minimal

0.53

 Stamp

0.53

Minimal

0.53

 Stamps

0.53

Hills

0.52

 hills

0.50

Activations Density 0.576%

references to places or geographical locations, particularly related to the term "Hills."

No Comments

No Known Activations