INDEX

Explanations

code-related syntax and structure, particularly focusing on attributes and method calls

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

RegressionTest

-0.71

ArrowToggle

-0.71

Efq

-0.68

 houſe

-0.67

aarrggbb

-0.66

StoreMessageInfo

-0.63

 Houſe

-0.63

 himſelf

-0.62

 myſelf

-0.61

 themſelves

-0.60

POSITIVE LOGITS

 true

0.65

true

0.64

 True

0.60

True

0.57

Datuak

0.55

 autorytatywna

0.54

False

0.54

 false

0.54

false

0.52

BOOLEAN

0.51

Activations Density 0.023%

code-related syntax and structure, particularly focusing on attributes and method calls

No Comments

No Known Activations