INDEX

Explanations

concepts and references related to training or educational programs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 Diocese

-0.83

 Osbourne

-0.83

 Antara

-0.82

 KeyCode

-0.81

Edd

-0.78

 Heard

-0.75

 Neel

-0.74

 spé

-0.72

 Sist

-0.72

 Spectroscopy

-0.71

POSITIVE LOGITS

 training

1.10

 trainings

0.98

 Training

0.96

 trainers

0.93

Training

0.91

 train

0.86

trained

0.86

training

0.85

 Trainings

0.84

TRAINING

0.84

Activations Density 0.049%

concepts and references related to training or educational programs

No Comments

No Known Activations