INDEX

Explanations

specified prefixes or words that are followed by another specific word or phrase

oai_token-act-pair · gpt-3.5-turbo

references to preferences and favorite things

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

xon

-0.94

CTV

-0.90

 Downs

-0.81

Trend

-0.79

tainment

-0.77

 Sakuya

-0.75

aneers

-0.73

 Canad

-0.70

 Guinness

-0.70

isons

-0.69

POSITIVE LOGITS

uler

0.95

ulously

0.83

 pref

0.79

penter

0.73

unci

0.69

opic

0.68

 alley

0.67

Ĺ

0.66

Obj

0.66

 debtor

0.65

Activations Density 0.030%

specified prefixes or words that are followed by another specific word or phrase

references to preferences and favorite things

No Comments

No Known Activations