INDEX

Explanations

questions, tips, answers, and instructional prompts

oai_token-act-pair · gpt-3.5-turbo

structured query and response formats in text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 disg

-0.76

cale

-0.70

itÃ©

-0.68

undai

-0.67

aps

-0.64

creen

-0.63

 indiscrim

-0.63

 glac

-0.62

paces

-0.61

ides

-0.60

POSITIVE LOGITS

0.99

 Number

0.98

 Yourself

0.96

 Summary

0.93

 Explan

0.89

 Description

0.86

!:

0.86

 Regarding

0.85

 Abuse

0.84

Activations Density 0.200%

questions, tips, answers, and instructional prompts

structured query and response formats in text

No Comments

No Known Activations