INDEX

Explanations

terms related to vaping and substances like kratom

oai_token-act-pair · gpt-3.5-turbo

references to specific products and substances, particularly e-cigarettes and kratom

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Rivers

-0.65

 Icar

-0.65

 Petty

-0.61

nm

-0.60

Merit

-0.59

UM

-0.58

rpm

-0.57

 USAF

-0.57

Uni

-0.57

UNC

-0.57

POSITIVE LOGITS

 proponents

0.82

 poses

0.72

 detractors

0.71

ervatives

0.70

 campaigners

0.70

 advocates

0.69

 skeptics

0.67

ervative

0.67

 scares

0.65

 negatives

0.65

Activations Density 0.357%

terms related to vaping and substances like kratom

references to specific products and substances, particularly e-cigarettes and kratom

No Comments

No Known Activations