INDEX

Explanations

terms related to a service or being of service

oai_token-act-pair · gpt-3.5-turbo

references to a service or the act of serving

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 heaviest

-0.63

 Kubrick

-0.63

 heavier

-0.62

 highlighting

-0.60

 prone

-0.60

 pregnant

-0.60

ãĥ£

-0.59

 prevailing

-0.59

orne

-0.58

 Donetsk

-0.58

POSITIVE LOGITS

iced

1.19

serv

1.07

itude

1.05

icing

1.01

iments

1.00

icer

0.89

ando

0.89

ecast

0.87

itary

0.87

estyles

0.85

Activations Density 0.022%

terms related to a service or being of service

references to a service or the act of serving

No Comments

No Known Activations