INDEX

Explanations

instances where something is being added or included

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the addition of new elements or enhancements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

grounds

-0.73

abe

-0.71

 engagements

-0.70

 meetings

-0.67

projects

-0.65

hots

-0.65

ongo

-0.64

results

-0.63

 searches

-0.63

POSITIVE LOGITS

 layer

1.21

 caveat

1.05

 dimension

1.03

 disclaimer

0.99

 twist

0.97

 notch

0.96

 element

0.86

 dime

0.85

 dash

0.83

 clause

0.82

Activations Density 0.159%

instances where something is being added or included

phrases that indicate the addition of new elements or enhancements

No Comments

No Known Activations