INDEX

Explanations

mentions of the word "pandemic."

oai_token-act-pair · gpt-3.5-turbo

terms related to medical conditions and diseases, particularly those ending in 'emic' and 'emia'

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

Not in Any Lists

No Comments

Negative Logits

SW

-0.75

aird

-0.73

CRE

-0.72

OW

-0.70

========

-0.68

 Aires

-0.68

Fill

-0.67

RECT

-0.67

adra

-0.65

Slot

-0.65

POSITIVE LOGITS

emic

1.23

onductor

0.84

soever

0.83

olon

0.80

 contag

0.78

ics

0.77

 Panic

0.76

mble

0.75

onduct

0.73

phal

0.73

Activations Density 0.011%

mentions of the word "pandemic."

terms related to medical conditions and diseases, particularly those ending in 'emic' and 'emia'

No Comments

No Known Activations