INDEX

Explanations

references to Canadian entities and topics

oai_token-act-pair · gpt-3.5-turbo

mentions of anything related to Canada

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

mble

-0.89

rex

-0.78

bler

-0.76

oldemort

-0.76

ppo

-0.76

iencies

-0.76

swick

-0.75

iets

-0.75

isen

-0.75

inse

-0.74

POSITIVE LOGITS

 Canadians

1.16

 Canadian

1.11

 Forces

1.09

istani

0.99

Canadian

0.98

 Centre

0.94

 Journalists

0.91

 Canad

0.90

 Scotia

0.87

 Rockies

0.86

Activations Density 0.009%

references to Canadian entities and topics

mentions of anything related to Canada

No Comments

No Known Activations