INDEX

Explanations

places or instances where actions are taken or formal requests are made

oai_token-act-pair · gpt-3.5-turbo

words related to lodging and accommodation

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

rawdownloadcloneembedreportprint

-0.70

enegger

-0.68

 Reaper

-0.67

onson

-0.67

schild

-0.63

 Saul

-0.63

wip

-0.62

vous

-0.61

 Cuomo

-0.59

Mub

-0.58

POSITIVE LOGITS

pole

1.28

itory

0.99

 lodge

0.91

nery

0.85

esta

0.83

der

0.83

holders

0.82

mate

0.82

holder

0.82

lie

0.78

Activations Density 0.017%

places or instances where actions are taken or formal requests are made

words related to lodging and accommodation

No Comments

No Known Activations