INDEX

Explanations

names of a person, particularly "Andrea"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Andrea" and variations of "rea"

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

creen

-0.87

paced

-0.86

doms

-0.86

lasses

-0.85

nown

-0.84

Ħ¢

-0.84

uilt

-0.83

mares

-0.82

quartered

-0.82

lif

-0.81

POSITIVE LOGITS

 Rossi

1.07

 Merkel

0.79

 Perez

0.78

 Andrea

0.76

Ces

0.75

 Mitchell

0.74

 Cros

0.73

 Grimes

0.72

 Morse

0.72

 Snyder

0.72

Activations Density 0.014%

names of a person, particularly "Andrea"

mentions of the name "Andrea" and variations of "rea"

No Comments

No Known Activations