INDEX

Explanations

references to legal terms and concepts related to liability and torts

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

-0.57

$_"

-0.55

httphttps

-0.54

 madonna

-0.53

 itſelf

-0.53

ɚ

-0.53

تفصیلات

-0.52

Autoritní

-0.52

replyreport

-0.52

ſelves

-0.51

POSITIVE LOGITS

 cases

0.39

law

0.39

TargetException

0.38



0.38

 jurí

0.36

 juridique

0.35

 Satz

0.35

葩

0.35

Law

0.34

 attempts

0.33

Activations Density 0.053%

references to legal terms and concepts related to liability and torts

No Comments

No Known Activations