INDEX

Explanations

specific patterns related to hierarchical structures or relationships in data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 disambiguazione

-0.66

帖最后由

-0.65

PerformLayout

-0.60

 propOrder

-0.60

ViewImports

-0.59

StoryboardSegue

-0.59

webElementXpaths

-0.57

 defaultstate

-0.57

MigrationBuilder

-0.56

 تضيفلها

-0.55

POSITIVE LOGITS

[toxicity=0]

0.39

 stället

0.38

ી

0.35

Autoritní

0.35

 године

0.34

 なっ

0.34

Â

0.33

Ή

0.32

########.

0.32

 gång

0.31

Activations Density 0.983%

specific patterns related to hierarchical structures or relationships in data

No Comments

No Known Activations