INDEX

Explanations

different modes or settings within a system or software

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

enegger

-0.84

mony

-0.79

istani

-0.77

ees

-0.74

areth

-0.73

ingly

-0.70

owan

-0.69

enda

-0.67

bones

-0.66

orian

-0.66

POSITIVE LOGITS

 selector

0.86

etting

0.81

 upgr

0.79

 Shift

0.77

 mode

0.72

 Modes

0.70

 activated

0.70

 whereby

0.67

 modes

0.67

 Mode

0.67

Activations Density 0.031%

different modes or settings within a system or software

No Comments

No Known Activations