INDEX

Explanations

references to physical spaces

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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

-0.72

IVERS

-0.68

 Turk

-0.63

ocide

-0.63

 Edmund

-0.61

 dism

-0.60

PUT

-0.60

 cartel

-0.60

 punishing

-0.59

 wrongful

-0.58

POSITIVE LOGITS

 spaces

1.26

hips

1.10

paces

1.03

 Spaces

1.02

hops

0.98

poons

0.98

pace

0.91

 gaps

0.90

uits

0.89

etter

0.88

Activations Density 0.009%

references to physical spaces

references to physical or conceptual spaces

No Comments

No Known Activations