INDEX

Explanations

mentions of the word "Miller"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Miller."

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

liest

-1.01

quo

-0.94

angular

-0.87

ulatory

-0.79

rous

-0.75

piring

-0.75

want

-0.74

lier

-0.73

liness

-0.73

ulates

-0.72

POSITIVE LOGITS

 Miller

1.00

 Lite

0.91

 Mayhem

0.83

hound

0.79

£ı

0.75

hawk

0.72

oche

0.71

beer

0.70

ophon

0.69

è¦ļéĨĴ

0.69

Activations Density 0.018%

mentions of the word "Miller"

mentions of the name "Miller."

No Comments

No Known Activations