INDEX

Explanations

questions followed by the word "Did"

oai_token-act-pair · gpt-3.5-turbo

the phrase "Did" in various contexts, indicating a focus on questions or inquiries

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

velop

-0.64

rooms

-0.64

rosis

-0.60

orse

-0.58

TeX

-0.57

berra

-0.57

Slot

-0.56

 range

-0.56

ogue

-0.56

BSD

-0.55

POSITIVE LOGITS

Did

3.33

Did

2.47

Was

1.95

 Didn

1.86

did

1.86

 Were

1.83

Was

1.69

 Does

1.68

DID

1.52

Had

1.49

Activations Density 0.013%

questions followed by the word "Did"

the phrase "Did" in various contexts, indicating a focus on questions or inquiries

No Comments

No Known Activations