INDEX

Explanations

quotations from sources

oai_token-act-pair · gpt-3.5-turbo

expressions that originate from or relate to specific individuals or sources

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

respective

-0.71

obyl

-0.66

required

-0.65

TEXT

-0.64

 accordingly

-0.63

 collateral

-0.63

devices

-0.63

mask

-0.62

film

-0.62

rina

-0.61

POSITIVE LOGITS

icial

0.86

 Rochester

0.79

 Alexandria

0.79

 Georgetown

0.78

 Honor

0.77

sky

0.77

 Anaheim

0.75

 Greater

0.75

 Bellev

0.74

ahu

0.74

Activations Density 0.057%

quotations from sources

expressions that originate from or relate to specific individuals or sources

No Comments

No Known Activations