INDEX

Explanations

references to insider trading

oai_token-act-pair · gpt-3.5-turbo

references to insider trading

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

Population

-0.73

ľ

-0.72

Pause

-0.67

interrupted

-0.67

 Parenthood

-0.66

 Archangel

-0.65

à¤

-0.65

enance

-0.65

ood

-0.64

ez

-0.63

POSITIVE LOGITS

 insider

0.95

 insiders

0.86

holder

0.79

obin

0.76

holders

0.75

aucus

0.74

 workings

0.72

clair

0.68

 trader

0.67

eteria

0.66

Activations Density 0.012%

references to insider trading

references to insider trading

No Comments

No Known Activations