INDEX

Explanations

technical terms related to programming attributes and data types

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.hook_resid_post

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

ItemBackground

-0.95

AsUp

-0.89

stateProvider

-0.78

windowFixed

-0.77

errHandler

-0.77

 kasarigan

-0.76

 unknownFields

-0.74

IVEREF

-0.73

ddots

-0.72

LookAnd

-0.71

POSITIVE LOGITS

Attribute

0.96

 Attribute

0.71

 attribute

0.69

CodeAttribute

0.60

setAttribute

0.59

adecimal

0.57

 PropTypes

0.55

pegno

0.54

 internetowa

0.54

sih

0.53

Activations Density 0.001%

technical terms related to programming attributes and data types

No Comments

No Known Activations