INDEX

Explanations

numbers related to time durations in minutes

oai_token-act-pair · gpt-3.5-turbo

mentions of time durations, specifically in minutes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.74

sonian

-0.70

spons

-0.68

uph

-0.66

 Moon

-0.65

 Bhar

-0.65

employ

-0.65

 Vest

-0.63

 Lunar

-0.62

enegger

-0.62

POSITIVE LOGITS

ftime

0.77

 stopp

0.76

 halftime

0.73

poons

0.71

 coached

0.71

 elapsed

0.69

Ago

0.67

 apiece

0.67

nesota

0.67

 Dodd

0.67

Activations Density 0.031%

numbers related to time durations in minutes

mentions of time durations, specifically in minutes

No Comments

No Known Activations