INDEX

Explanations

mention of various crimes and legal terms related to theft and property damage

oai_token-act-pair · gpt-3.5-turbo

mentions of property and property-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ACTED

-0.80

 Modest

-0.75

arios

-0.74

ursday

-0.74

ansen

-0.70

ature

-0.67

zee

-0.66

Bie

-0.66

ould

-0.65

yo

-0.64

POSITIVE LOGITS

 ownership

0.95

 owners

0.93

owner

0.93

Property

0.88

 holdings

0.87

owners

0.87

 property

0.86

 rights

0.85

 owner

0.82

property

0.79

Activations Density 0.010%

mention of various crimes and legal terms related to theft and property damage

mentions of property and property-related concepts

No Comments

No Known Activations