INDEX

Explanations

names of professions, particularly lawyers

oai_token-act-pair · gpt-3.5-turbo

references to legal professionals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

oppers

-0.76

plings

-0.73

 disembark

-0.68

 Reincarn

-0.66

Poll

-0.66

bows

-0.65

 Flavoring

-0.63

phal

-0.63

 Railway

-0.62

 Yards

-0.62

POSITIVE LOGITS

yer

0.90

 warr

0.86

yers

0.83

suit

0.82

maid

0.81

 solicitor

0.79

 Gloria

0.77

suits

0.77

 examiner

0.77

 specializing

0.71

Activations Density 0.039%

names of professions, particularly lawyers

references to legal professionals

No Comments

No Known Activations