INDEX

Explanations

religious terminology or discussions related to religion

oai_token-act-pair · gpt-3.5-turbo

references to religion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

-0.76

 Lans

-0.70

msg

-0.69

Yon

-0.69

astern

-0.68

sg

-0.67

berry

-0.66

amaz

-0.66

ptives

-0.65

 Packers

-0.64

POSITIVE LOGITS

 religion

0.93

ophobia

0.84

 worshipped

0.80

 affiliation

0.80

 scriptures

0.78

zai

0.75

 fundamentalist

0.74

antry

0.73

ophobic

0.73

igion

0.73

Activations Density 0.009%

religious terminology or discussions related to religion

references to religion

No Comments

No Known Activations