INDEX

Explanations

time durations expressed in hours

oai_token-act-pair · gpt-3.5-turbo

references to durations of time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

LECT

-0.70

Cole

-0.69

esan

-0.68

Shock

-0.66

chance

-0.65

founded

-0.65

vention

-0.65

Sov

-0.63

ven

-0.62

disc

-0.62

POSITIVE LOGITS

 Hours

1.02

Ago

0.97

glass

0.95

 hours

0.95

hrs

0.93

endment

0.93

mith

0.91

Hours

0.88

 elapsed

0.85

pring

0.82

Activations Density 0.018%

time durations expressed in hours

references to durations of time

No Comments

No Known Activations