INDEX

Explanations

numerical data such as quantities and measurements

oai_token-act-pair · gpt-3.5-turbo

numeric data and quantitative expressions related to events or activities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Wife

-0.85

word

-0.74

someone

-0.72

sense

-0.71

translation

-0.70

ONEY

-0.70

 Supplement

-0.69

agonist

-0.69

ebus

-0.68

 rider

-0.67

POSITIVE LOGITS

 locations

1.82

 sites

1.78

 residences

1.72

 venues

1.67

 warehouses

1.62

 campuses

1.61

 facilities

1.61

 buildings

1.59

 establishments

1.59

 malls

1.59

Activations Density 0.516%

numerical data such as quantities and measurements

numeric data and quantitative expressions related to events or activities

No Comments

No Known Activations