INDEX

Explanations

technical terms related to machine learning model handling and errors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 olvidado

-0.43

 geçi

-0.42

 asesinado

-0.41

 sprá

-0.41

ustimmung

-0.40

 AssemblyTitle

-0.40

ConstraintMaker

-0.38

 présents

-0.37

Gön

-0.36

nexo

-0.36

POSITIVE LOGITS

 training

0.79

 neural

0.74

 tensor

0.70

 Tensor

0.69

 trained

0.68

 torch

0.68

 Training

0.68

 model

0.67

 train

0.65

 Neural

0.65

Activations Density 0.599%

technical terms related to machine learning model handling and errors

No Comments

No Known Activations