INDEX

Explanations

the word "Nothing" at a high activation level in sentences

oai_token-act-pair · gpt-3.5-turbo

repetitive expressions of the word "Nothing" implying a lack of significant change or impact

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

 households

-0.74

 estimate

-0.66

 estimates

-0.65

alian

-0.63

 host

-0.62

 over

-0.61

 estimation

-0.60

 families

-0.60

 ranking

-0.60

 advantage

-0.60

POSITIVE LOGITS

 Nothing

3.30

Nothing

2.93

nothing

2.30

 nothing

1.88

 Anything

1.66

 Something

1.65

 Everything

1.60

 None

1.59

 Nobody

1.38

None

1.36

Activations Density 0.017%

the word "Nothing" at a high activation level in sentences

repetitive expressions of the word "Nothing" implying a lack of significant change or impact

No Comments

No Known Activations