INDEX

Explanations

mentions of individuals being trained for specific tasks or roles

oai_token-act-pair · gpt-3.5-turbo

references to individuals or groups that are specifically trained for various professions or tasks

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

FACE

-0.77

 goodbye

-0.67

pler

-0.67

 Proposition

-0.65

adish

-0.65

plet

-0.64

hazard

-0.63

apest

-0.63

issues

-0.62

bish

-0.62

POSITIVE LOGITS

arnaev

0.74

glers

0.73

 trained

0.73

erto

0.73

 certific

0.72

anamo

0.72

romy

0.71

 athletes

0.69

 instructors

0.69

 professionals

0.68

Activations Density 0.036%

mentions of individuals being trained for specific tasks or roles

references to individuals or groups that are specifically trained for various professions or tasks

No Comments

No Known Activations