INDEX

Explanations

quotations and indirect speech patterns

oai_token-act-pair · gpt-3.5-turbo

direct quotes from individuals in a news context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 envy

-0.93

 egalitarian

-0.87

 utopian

-0.87

 feminism

-0.84

 champion

-0.84

 muse

-0.83

 philosophers

-0.83

 embracing

-0.83

 critiques

-0.81

 utilitarian

-0.80

POSITIVE LOGITS

Officers

1.27

Initial

1.10

Investigators

1.10

Officials

1.08

Emergency

1.04

Staff

1.04

Police

1.03

iminary

1.00

Authorities

0.98

roximately

0.96

Activations Density 0.413%

quotations and indirect speech patterns

direct quotes from individuals in a news context

No Comments

No Known Activations