INDEX

Explanations

numbers and locations in texts

oai_token-act-pair · gpt-3.5-turbo

instances of significant numerics related to locations and transformations

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

videos

-0.64

soDeliveryDate

-0.64

[+

-0.62

References

-0.61

Conclusion

-0.61

 CrossRef

-0.61

clinton

-0.60

soType

-0.60

itutes

-0.58

 Regarding

-0.58

POSITIVE LOGITS

 sprawling

1.12

 towering

1.02

 enigmatic

0.95

 tiny

0.92

 sleek

0.91

 notoriously

0.91

 hugely

0.91

 elusive

0.90

 coveted

0.89

 relentless

0.87

Activations Density 1.019%

numbers and locations in texts

instances of significant numerics related to locations and transformations

No Comments

No Known Activations