INDEX

Explanations

mentions of the quality of various things

oai_token-act-pair · gpt-3.5-turbo

terms related to quality

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

opa

-0.78

erald

-0.71

coni

-0.70

stall

-0.70

vention

-0.68

canon

-0.67

aida

-0.67

ften

-0.65

ker

-0.65

cart

-0.64

POSITIVE LOGITS

 assurance

1.32

 quality

1.01

Quality

0.91

 improvement

0.87

Reviewer

0.80

 Quality

0.79

 retention

0.74

quality

0.74

 Emin

0.73

 metrics

0.73

Activations Density 0.019%

mentions of the quality of various things

terms related to quality

No Comments

No Known Activations