INDEX

Explanations

phrases related to limits and maximums

oai_token-act-pair · gpt-3.5-turbo

phrases indicating limits, quantities, or restrictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

pse

-0.76

 Zimmer

-0.74

}}}

-0.66

liest

-0.65

ãĤ¨

-0.63

 Assy

-0.62

å®

-0.60

âĹ¼

-0.60

ãĤ¿

-0.59

Redd

-0.59

POSITIVE LOGITS

 unlimited

0.90

maximum

0.86

 prepaid

0.85

 increments

0.82

 drinks

0.80

 servings

0.80

abyte

0.78

 maximum

0.78

 Maximum

0.77

each

0.74

Activations Density 0.456%

phrases related to limits and maximums

phrases indicating limits, quantities, or restrictions

No Comments

No Known Activations