INDEX

Explanations

formatted code structure elements, particularly focusing on metadata and configurations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

 Shakspeare

-1.03

Efq

-1.03

 Monfieur

-1.00

 للمعارف

-0.98

IsContent

-0.93

 Majefty

-0.90

 Jefus

-0.89

 poffible

-0.89

 ―――――

-0.88

MemoryWarning

-0.88

POSITIVE LOGITS

0.79

BeginContext

0.76

EndContext

0.76

“

0.68

cu

0.61

 Bell

0.56

man

0.54

0.53

0.52

roll

0.52

Activations Density 0.007%

formatted code structure elements, particularly focusing on metadata and configurations

No Comments

No Known Activations