INDEX

Explanations

positive things or phrases

oai_token-act-pair · gpt-3.5-turbo

expressions of positive or supportive sentiments

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

atars

-0.80

otom

-0.80

ĸļ

-0.78

umb

-0.70

 guiActiveUn

-0.70

pora

-0.63

irez

-0.63

 phases

-0.62

igham

-0.61

 veins

-0.60

POSITIVE LOGITS

nered

0.78

mire

0.70

tein

0.70

manship

0.69

 outweigh

0.69

 Mahjong

0.69

ilda

0.65

âĹ¼

0.64

 luck

0.64

standing

0.62

Activations Density 0.710%

positive things or phrases

expressions of positive or supportive sentiments

No Comments

No Known Activations