INDEX

Explanations

actions related to starting, advancing, or completing tasks or projects

oai_token-act-pair · gpt-3.5-turbo

phrases related to initiating and progressing tasks or projects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 exceptions

-0.69

 hint

-0.65

 exception

-0.63

 fractions

-0.63

 href

-0.61

DragonMagazine

-0.61

 recourse

-0.60

 pleading

-0.60

ettings

-0.60

 betray

-0.58

POSITIVE LOGITS

 smoothly

0.90

 ASAP

0.85

 efficiently

0.81

 cheaply

0.76

 properly

0.76

 safely

0.70

 quicker

0.70

 unim

0.69

 sooner

0.68

washer

0.67

Activations Density 0.334%

actions related to starting, advancing, or completing tasks or projects

phrases related to initiating and progressing tasks or projects

No Comments

No Known Activations