INDEX

Explanations

patterns related to file paths and directory structures in programming contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Early

-0.54

Early

-0.52

dotte

-0.52

dată

-0.51

Extinguishing

-0.50

 szükség

-0.49

xically

-0.48

を含む

-0.48

녁

-0.48

 dente

-0.47

POSITIVE LOGITS

SBATCH

0.83

<bos>

0.82

Datuak

0.77

Clik

0.75

TagMode

0.69

'{@

0.67

 EconPapers

0.67

Portail

0.66

bootstrapcdn

0.66

getSystemService

0.65

Activations Density 0.059%

patterns related to file paths and directory structures in programming contexts

No Comments

No Known Activations