INDEX

Explanations

references to the country "Canada."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Canada."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bler

-0.82

sym

-0.77

lihood

-0.70

inse

-0.68

tle

-0.67

eer

-0.66

terior

-0.66

chest

-0.66

blers

-0.65

angler

-0.65

POSITIVE LOGITS

 Goose

1.13

 Tire

1.02

 Centre

0.90

 Scotia

0.88

 Institutes

0.87

 Revenue

0.85

 Forces

0.85

Canada

0.83

 Geographic

0.82

 Canadians

0.82

Activations Density 0.045%

references to the country "Canada."

occurrences of the word "Canada."

No Comments

No Known Activations