INDEX

Explanations

mentions of promises or commitments

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "promise," particularly in contexts involving commitments or expectations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 quarters

-0.71

 Lauderdale

-0.68

agram

-0.64

 Occupations

-0.64

icult

-0.62

Center

-0.61

parts

-0.61

 viewed

-0.61

icial

-0.61

apsed

-0.60

POSITIVE LOGITS

 promise

4.22

 promises

2.64

 Promise

2.48

 promised

1.82

 promising

1.79

 pledge

1.76

 guarantee

1.70

vow

1.69

 hope

1.41

 pledges

1.37

Activations Density 0.007%

mentions of promises or commitments

references to the concept of "promise," particularly in contexts involving commitments or expectations

No Comments

No Known Activations