INDEX

Explanations

statements where there is agreement or consensus

oai_token-act-pair · gpt-3.5-turbo

statements of consensus or agreement

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

oufl

-0.69

 Ascension

-0.65

ADS

-0.64

illus

-0.64

zh

-0.64

DAQ

-0.63

 Advent

-0.63

ãĤ¹

-0.62

ember

-0.62

dimension

-0.61

POSITIVE LOGITS

 agrees

1.08

agree

0.99

 agree

0.91

reement

0.85

ilibrium

0.84

terday

0.83

ajor

0.81

 agreeing

0.80

reements

0.79

 agreed

0.78

Activations Density 0.003%

statements where there is agreement or consensus

statements of consensus or agreement

No Comments

No Known Activations