INDEX

Explanations

days of the week

oai_token-act-pair · gpt-3.5-turbo

mentions of specific days of the week, particularly Sunday

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

reprene

-0.95

CHQ

-0.80

chwitz

-0.80

commun

-0.79

itarian

-0.79

pora

-0.75

yll

-0.74

ensual

-0.74

=-=-=-=-=-=-=-=-

-0.74

ongs

-0.73

POSITIVE LOGITS

 scrimmage

1.07

 matchup

1.01

 afternoon

1.00

 season

1.00

 halftime

1.00

 preseason

0.98

 Stadium

0.95

 injury

0.95

MVP

0.93

 playoffs

0.92

Activations Density 0.138%

days of the week

mentions of specific days of the week, particularly Sunday

No Comments

No Known Activations