INDEX

Explanations

political terms related to partisanship or political loyalty

oai_token-act-pair · gpt-3.5-turbo

terms related to political affiliations and relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Scotia

-0.71

ASED

-0.64

bye

-0.64

 Slater

-0.63

 spree

-0.61

 marrow

-0.61

 Wilde

-0.60

 Pearce

-0.60

 Breach

-0.60

Osc

-0.60

POSITIVE LOGITS

hip

2.40

hips

1.99

ial

1.48

edIn

1.25

ions

1.17

ially

1.13

alty

1.12

uit

1.12

ships

1.11

ation

1.11

Activations Density 0.091%

political terms related to partisanship or political loyalty

terms related to political affiliations and relationships

No Comments

No Known Activations