INDEX

Explanations

words that are used in a broad or general sense

oai_token-act-pair · gpt-3.5-turbo

phrases that use the term "broadly" to indicate generalizations or overarching concepts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eon

-0.82

 Million

-0.74

 Tycoon

-0.74

 Rebels

-0.67

 Thousand

-0.66

 Keeper

-0.66

ÄŁ

-0.66

 Gorge

-0.66

 Camer

-0.65

tein

-0.63

POSITIVE LOGITS

soType

1.04

sheet

0.98

 defined

0.93

 construed

0.91

atform

0.84

ensical

0.84

baum

0.84

sheets

0.83

 speaking

0.80

 categorized

0.78

Activations Density 0.005%

words that are used in a broad or general sense

phrases that use the term "broadly" to indicate generalizations or overarching concepts

No Comments

No Known Activations