INDEX

Explanations

points in a text where the speaker or writer continues a statement or idea

oai_token-act-pair · gpt-3.5-turbo

statements indicating continuation or elaboration in dialogue

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

MpServer

-0.67

Born

-0.66

knit

-0.64

surv

-0.62

soDeliveryDate

-0.60

aughtered

-0.59

brance

-0.59

mite

-0.57

 civilisation

-0.56

usercontent

-0.55

POSITIVE LOGITS

 sarcast

1.14

 rhet

0.93

 quoting

0.82

 reiter

0.78

 noting

0.77

 reiterate

0.77

 emphasizing

0.76

:]

0.75

 emphatically

0.75

 mentioning

0.74

Activations Density 0.727%

points in a text where the speaker or writer continues a statement or idea

statements indicating continuation or elaboration in dialogue

No Comments

No Known Activations