INDEX

Explanations

years represented in a specific format

oai_token-act-pair · gpt-3.5-turbo

references to historical periods, particularly those indicated by specific years

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

iants

-0.88

aminer

-0.84

pora

-0.76

iant

-0.76

phia

-0.75

utenberg

-0.70

afort

-0.70

oaded

-0.69

arious

-0.69

plings

-0.67

POSITIVE LOGITS

0.98

âĸĪâĸĪ

0.97

0.89

0.86

0.85

0.84

Activations Density 0.014%

years represented in a specific format

references to historical periods, particularly those indicated by specific years

No Known Activations