INDEX

Explanations

time indicators denoting the future

oai_token-act-pair · gpt-3.5-turbo

references to upcoming events or timelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

baugh

-0.76

hess

-0.75

ocker

-0.66

lees

-0.65

 Feldman

-0.64

had

-0.63

ordes

-0.62

cks

-0.62

 Helic

-0.62

lee

-0.60

POSITIVE LOGITS

 week

1.28

 month

1.26

 year

1.17

decade

0.96

 semester

0.95

 generation

0.92

 fortnight

0.91

yr

0.89

 spring

0.88

ĻĤ

0.86

Activations Density 0.032%

time indicators denoting the future

references to upcoming events or timelines

No Comments

No Known Activations