INDEX

Explanations

mentions of specific durations of time in hours

oai_token-act-pair · gpt-3.5-turbo

references to durations of time measured in hours

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

raised

-0.68

osate

-0.67

govern

-0.67

raud

-0.66

rez

-0.64

forward

-0.64

bender

-0.63

bred

-0.61

rup

-0.60

crop

-0.60

POSITIVE LOGITS

 hours

3.78

 Hours

2.86

hours

2.73

hrs

2.62

Hours

2.53

 hour

2.34

hour

2.13

 minutes

2.05

 days

1.98

 weeks

1.88

Activations Density 0.021%

mentions of specific durations of time in hours

references to durations of time measured in hours

No Comments

No Known Activations