INDEX

Explanations

time-related terms, specifically durations in months

oai_token-act-pair · gpt-3.5-turbo

references to time spans or durations, specifically related to months

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

oka

-0.73

enter

-0.70

heid

-0.68

iva

-0.65

hess

-0.61

Original

-0.55

cussion

-0.55

 conserve

-0.55

ius

-0.55

lich

-0.55

POSITIVE LOGITS

 months

3.44

 weeks

2.81

months

2.79

 Months

2.74

 years

2.29

 days

2.21

month

1.92

 month

1.90

years

1.90

 hours

1.85

Activations Density 0.026%

time-related terms, specifically durations in months

references to time spans or durations, specifically related to months

No Comments

No Known Activations