INDEX

Explanations

phrases related to impressiveness or extreme positivity

oai_token-act-pair · gpt-3.5-turbo

strong emotional descriptions and superlatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Caption

-0.81

 grievances

-0.73

>[

-0.70

 surn

-0.65

 Partial

-0.64

ND

-0.62

 coun

-0.62

 relinqu

-0.60

ida

-0.60

 surname

-0.60

POSITIVE LOGITS

ciating

1.25

 unbeliev

0.99

 pandemonium

0.91

 phenomenal

0.91

 incredible

0.88

 Awesome

0.84

 unbelievable

0.83

ĸļ

0.83

 amazing

0.83

 amaz

0.82

Activations Density 0.620%

phrases related to impressiveness or extreme positivity

strong emotional descriptions and superlatives

No Comments

No Known Activations