INDEX

Explanations

technology-related terms, especially those related to interconnected devices

oai_token-act-pair · gpt-3.5-turbo

mentions of the phrase "Things" or variations of "Thing."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

onym

-0.73

ctic

-0.73

 anonymity

-0.72

oufl

-0.71

otto

-0.71

ECD

-0.70

por

-0.68

 fres

-0.67

idated

-0.66

pas

-0.66

POSITIVE LOGITS

 Things

1.08

Things

1.05

 Happ

0.99

 Matters

0.91

 Bullets

0.81

 Events

0.80

 Worse

0.80

 Lies

0.76

 Affect

0.75

 Valiant

0.73

Activations Density 0.022%

technology-related terms, especially those related to interconnected devices

mentions of the phrase "Things" or variations of "Thing."

No Comments

No Known Activations