INDEX

Explanations

phrases that indicate a deviation from the norm or regularity

oai_token-act-pair · gpt-3.5-turbo

phrases indicating normal or typical behaviors or situations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Kob

-0.68

Kut

-0.67

ged

-0.61

addons

-0.61

yet

-0.60

 Bern

-0.59

 Universe

-0.59

 Kush

-0.59

Hung

-0.59

health

-0.58

POSITIVE LOGITS

 disclaim

0.90

entimes

0.90

 suspects

0.87

 reserved

0.85

 aspir

0.78

 refers

0.76

 regarded

0.75

 speaking

0.74

 relegated

0.73

 associated

0.73

Activations Density 0.055%

phrases that indicate a deviation from the norm or regularity

phrases indicating normal or typical behaviors or situations

No Comments

No Known Activations