INDEX

Explanations

words related to physical spaces or locations

oai_token-act-pair · gpt-3.5-turbo

references to physical or conceptual spaces

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 vengeance

-0.73

 Turk

-0.68

IVERS

-0.66

 Rupert

-0.64

ocide

-0.63

 blackmail

-0.62

 neurot

-0.62

POV

-0.62

 dism

-0.62

ctive

-0.60

POSITIVE LOGITS

 spaces

1.20

hips

1.18

 Spaces

0.99

hops

0.96

uits

0.95

paces

0.91

pace

0.87

dayName

0.86

lot

0.85

poons

0.85

Activations Density 0.015%

words related to physical spaces or locations

references to physical or conceptual spaces

No Comments

No Known Activations