INDEX

Explanations

instances of the word "stock"

oai_token-act-pair · gpt-3.5-turbo

references to stock and stock market performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Homeless

-0.77

Lod

-0.67

 Dough

-0.66

advertisement

-0.64

patient

-0.63

soever

-0.63

 Mandatory

-0.62

inx

-0.61

 Bethesda

-0.61

 Archdemon

-0.60

POSITIVE LOGITS

holders

0.91

 stocks

0.87

stock

0.84

stocks

0.84

seed

0.83

gage

0.83

 stock

0.83

holm

0.79

roller

0.77

holder

0.76

Activations Density 0.008%

instances of the word "stock"

references to stock and stock market performance

No Comments

No Known Activations