INDEX

Explanations

phrases related to significant events, controversies, and actions

oai_token-act-pair · gpt-3.5-turbo

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Quantity

-0.85

 Awakens

-0.79

 Inventory

-0.72

 Narc

-0.71

 Instruction

-0.69

 warranties

-0.69

 Installation

-0.67

olution

-0.67

 Explicit

-0.66

####

-0.65

POSITIVE LOGITS

 criticised

1.36

 criticized

1.33

 denounced

1.25

 hotly

1.20

 condemned

1.18

 ridiculed

1.17

 greeted

1.16

 cheered

1.16

 hailed

1.16

 applauded

1.15

Activations Density 0.143%

phrases related to significant events, controversies, and actions

No Comments

No Known Activations