INDEX

Explanations

phrases and sentences containing words described with positive attributes

oai_token-act-pair · gpt-3.5-turbo

positive descriptors, particularly the word "brilliant."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

avis

-0.75

eper

-0.72

intestinal

-0.72

asta

-0.69

©¶æ

-0.68

VR

-0.68

VI

-0.67

borg

-0.67

alone

-0.67

aver

-0.67

POSITIVE LOGITS

 brilliance

0.82

ulously

0.82

 minds

0.79

 brilliant

0.79

 Brilliant

0.76

sonian

0.74

 rainbow

0.72

 genius

0.72

NESS

0.70

minded

0.69

Activations Density 0.008%

phrases and sentences containing words described with positive attributes

positive descriptors, particularly the word "brilliant."

No Comments

No Known Activations