INDEX

Explanations

references to the name "Jack" and technical terms related to software patches

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

Eloquent

-0.47

Hochspringen

-0.43

ismet

-0.43

sizeCache

-0.40

IsContent

-0.39

 viewType

-0.39

 pres

-0.39

 Steiner

-0.38

 Wayback

-0.37

 esophagus

-0.36

POSITIVE LOGITS

patch

0.79

Patch

0.71

 patch

0.70

 Patch

0.67

 PATCH

0.66

PATCH

0.64

bug

0.63

Jack

0.62

 Jack

0.61

 chips

0.61

Activations Density 0.187%

references to the name "Jack" and technical terms related to software patches

No Comments

No Known Activations