INDEX

Explanations

phrases related to rules, agreements, or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases related to knowledge and personal growth

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

bilt

-0.68

reditary

-0.62

quished

-0.61

 Hels

-0.60

ibaba

-0.60

neau

-0.60

ife

-0.59

.?

-0.59

chin

-0.57

 Cosponsors

-0.57

POSITIVE LOGITS

 hindsight

0.66

 nobody

0.64

 totality

0.63

cases

0.62

pires

0.60

 consciousness

0.58

 persuasion

0.57

])

0.56

 chances

0.56

 intercourse

0.56

Activations Density 0.582%

phrases related to rules, agreements, or concepts

phrases related to knowledge and personal growth

No Comments

No Known Activations