INDEX

Explanations

references to optics and optical technology, specifically associated with the term "Opt"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "Opt" and its associated terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ptives

-0.71

FORE

-0.70

UGH

-0.70

ãĥĨãĤ£

-0.70

Madison

-0.66

INESS

-0.65

è¦ļéĨĴ

-0.64

URRENT

-0.64

ppo

-0.63

 withd

-0.63

POSITIVE LOGITS

imum

1.24

imal

1.04

icians

0.96

ician

0.93

ima

0.92

ometry

0.91

imus

0.87

ics

0.85

uary

0.84

aye

0.83

Activations Density 0.008%

references to optics and optical technology, specifically associated with the term "Opt"

variations of the word "Opt" and its associated terms

No Comments

No Known Activations