INDEX

Explanations

words related to rules and regulations

oai_token-act-pair · gpt-3.5-turbo

references to specific regulations or rules, particularly "Rule 9" and "Rule 8."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

itate

-0.86

acters

-0.82

itÃ©

-0.82

 Pradesh

-0.75

 Hots

-0.73

itant

-0.72

apest

-0.69

izoph

-0.68

imar

-0.68

velength

-0.68

POSITIVE LOGITS

book

1.28

books

1.14

making

1.00

breakers

0.95

breaker

0.94

maker

0.93

makers

0.90

witz

0.85

 breaker

0.83

Rule

0.80

Activations Density 0.025%

words related to rules and regulations

references to specific regulations or rules, particularly "Rule 9" and "Rule 8."

No Comments

No Known Activations