INDEX

Explanations

mentions of ex-partners or ex-relationships

oai_token-act-pair · gpt-3.5-turbo

references to previous relationships or partners

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 GOODMAN

-0.76

 Gork

-0.65

 baskets

-0.64

 cones

-0.64

 halls

-0.64

 Inspection

-0.62

 revolving

-0.61

 similarities

-0.60

 Flavoring

-0.60

 disse

-0.59

POSITIVE LOGITS

clamation

1.33

uber

1.24

orbit

1.11

communication

1.09

girlfriend

1.08

clave

1.04

oplan

1.02

cell

1.02

ogenous

1.00

terior

1.00

Activations Density 0.016%

mentions of ex-partners or ex-relationships

references to previous relationships or partners

No Comments

No Known Activations