INDEX

Explanations

mentions of the word "Valentine" at various intensities

oai_token-act-pair · gpt-3.5-turbo

mentions of Valentine's Day and related terms

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

ulhu

-0.99

alach

-0.91

lopp

-0.86

ologies

-0.79

atche

-0.75

rote

-0.72

wered

-0.72

artment

-0.72

ologically

-0.71

raltar

-0.70

POSITIVE LOGITS

 Valentine

0.93

 Staples

0.84

 Valent

0.72

 Claus

0.69

 Jarrett

0.67

 Vigil

0.63

 Payne

0.62

 Weekend

0.62

stice

0.62

 Sprite

0.62

Activations Density 0.011%

mentions of the word "Valentine" at various intensities

mentions of Valentine's Day and related terms

No Comments

No Known Activations