INDEX

Explanations

references to religious figures, particularly priests

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "priest" and associated contexts

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

DonaldTrump

-0.82

BAT

-0.74

 Flavoring

-0.68

oday

-0.67

RFC

-0.66

OX

-0.63

TING

-0.63

issors

-0.63

ilater

-0.63

ggles

-0.62

POSITIVE LOGITS

esses

1.59

ess

1.11

 priests

1.05

 priest

0.96

osate

0.88

 ordained

0.87

mares

0.78

ificial

0.77

hood

0.75

 exorc

0.74

Activations Density 0.014%

references to religious figures, particularly priests

mentions of the term "priest" and associated contexts

No Comments

No Known Activations