INDEX

Explanations

positions or roles of authority

oai_token-act-pair · gpt-3.5-turbo

phrases related to various positions, roles, and titles held by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raq

-0.75

jad

-0.74

itions

-0.74

NetMessage

-0.73

 reperto

-0.73

kai

-0.72

urrencies

-0.71

ples

-0.69

Apps

-0.69

venants

-0.68

POSITIVE LOGITS

 protector

0.90

 overse

0.87

 commander

0.86

 medi

0.86

 ambassador

0.85

 presiding

0.83

 lieutenant

0.82

assador

0.81

 sergeant

0.81

 councillor

0.79

Activations Density 0.163%

positions or roles of authority

phrases related to various positions, roles, and titles held by individuals

No Comments

No Known Activations