INDEX

Explanations

explicit statements or assertions in a text

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate reporting or referencing information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 conquered

-0.66

 Acquisition

-0.65

 downed

-0.65

 fitt

-0.63

 pandemonium

-0.62

 contracted

-0.61

 Patron

-0.60

 Legendary

-0.60

 Cooldown

-0.59

 Frie

-0.59

POSITIVE LOGITS

 mention

0.94

 assertions

0.84

uggest

0.84

 eluc

0.83

 debunk

0.81

reference

0.81

 summarize

0.81

 explicitly

0.80

 conclusions

0.80

 mentions

0.80

Activations Density 0.224%

explicit statements or assertions in a text

phrases that indicate reporting or referencing information

No Comments

No Known Activations