INDEX

Explanations

personal emotional responses related to oneself and others

oai_token-act-pair · gpt-3.5-turbo

expressions of personal emotional experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

foreseen

-0.79

accompan

-0.78

EngineDebug

-0.70

 occurring

-0.66

itiz

-0.66

tesy

-0.65

ItemTracker

-0.64

Ore

-0.64

 accrued

-0.63

MpServer

-0.63

POSITIVE LOGITS

 rethink

1.29

 realise

1.24

 feel

1.23

 realize

1.22

 cringe

1.20

 blush

1.17

 jealous

1.16

 laugh

1.14

 uncomfortable

1.14

 ashamed

1.11

Activations Density 0.054%

personal emotional responses related to oneself and others

expressions of personal emotional experiences

No Comments

No Known Activations