INDEX

Explanations

time-related words and phrases, particularly those indicating various durations such as months, years, and decades

oai_token-act-pair · gpt-3.5-turbo

references to time durations, particularly years and months

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rone

-0.68

ople

-0.66

itely

-0.63

ocaly

-0.63

nels

-0.63

adium

-0.62

roit

-0.60

anooga

-0.60

ibrary

-0.59

jac

-0.59

POSITIVE LOGITS

ago

1.53

Ago

1.30

 elapsed

1.05

 later

0.97

ago

0.96

 transpired

0.90

 hindsight

0.84

 earlier

0.83

shy

0.82

 passed

0.80

Activations Density 0.070%

time-related words and phrases, particularly those indicating various durations such as months, years, and decades

references to time durations, particularly years and months

No Comments

No Known Activations