INDEX

Explanations

code snippets and related programming concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 Monfieur

-0.79

 itſelf

-0.78

 Majefty

-0.76

 myſelf

-0.75

 themſelves

-0.73

 esternos

-0.72

 Houſe

-0.72

 Verſ

-0.71

 OFDb

-0.69

ReusableCell

-0.69

POSITIVE LOGITS

 tried

0.44

try

0.44

try

0.40

ad

0.40

0.39

 tries

0.35

 erfol

0.35

si

0.35

ze

0.34

Activations Density 0.256%

code snippets and related programming concepts

No Comments

No Known Activations