INDEX

Explanations

the term "Rift" specifically

oai_token-act-pair · gpt-3.5-turbo

mentions of the Oculus Rift virtual reality headset

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 abbrevi

-0.77

living

-0.66

å¥³

-0.64

 Dodd

-0.64

hours

-0.62

compl

-0.62

 revolving

-0.61

ARDS

-0.60

effective

-0.58

 bureaucratic

-0.58

POSITIVE LOGITS

 Rift

1.07

DK

0.94

 headset

0.90

 Rivals

0.84

 Oculus

0.80

idge

0.79

izont

0.79

ï¸

0.79

 headsets

0.79

asonic

0.75

Activations Density 0.010%

the term "Rift" specifically

mentions of the Oculus Rift virtual reality headset

No Comments

No Known Activations