INDEX

Explanations

quantities of items or objects

oai_token-act-pair · gpt-3.5-turbo

mentions of "units" in various contexts, particularly relating to sales and housing

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

é»Ĵ

-0.73

vous

-0.69

 havoc

-0.68

 vengeance

-0.67

ãĤī

-0.64

uez

-0.63

æĢ

-0.63

lder

-0.61

 Roberts

-0.61

romeda

-0.61

POSITIVE LOGITS

 units

1.22

units

1.07

 Units

0.98

 unit

0.94

imet

0.91

matically

0.85

Unit

0.84

unit

0.83

 Unit

0.83

mates

0.82

Activations Density 0.009%

quantities of items or objects

mentions of "units" in various contexts, particularly relating to sales and housing

No Comments

No Known Activations