INDEX

Explanations

language related to addiction

oai_token-act-pair · gpt-3.5-turbo

references to addiction in various contexts

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

erial

-0.75

Kov

-0.75

vard

-0.68

prints

-0.67

FIR

-0.66

CFR

-0.65

eor

-0.65

 Collider

-0.64

 Rouge

-0.64

FAA

-0.63

POSITIVE LOGITS

 addiction

1.10

 relapse

0.99

 addict

0.93

 addicted

0.91

 Addiction

0.87

olesc

0.86

encies

0.85

 craving

0.83

 rehab

0.82

 proble

0.82

Activations Density 0.019%

language related to addiction

references to addiction in various contexts

No Comments

No Known Activations