INDEX

Explanations

references to legal concepts and terminology

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

asInstanceOf

-0.49

Amit

-0.49

 Gibbons

-0.49

Mit

-0.48

di

-0.48

 Whittaker

-0.48

Vicente

-0.47

 McIntyre

-0.47

indeki

-0.47

 Morrissey

-0.47

POSITIVE LOGITS

law

1.20

LAW

1.16

Law

1.14

Law

1.04

law

0.98

LAW

0.97

 laws

0.97

 Laws

0.96

Laws

0.94

 LAWS

0.94

Activations Density 0.026%

references to legal concepts and terminology

No Comments

No Known Activations