INDEX

Explanations

special characters and formatting related to text adjustments

oai_token-act-pair · gpt-3.5-turbo

symbols or characters that signify special formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

alty

-0.83

urgy

-0.77

cliffe

-0.75

GHz

-0.72

anguage

-0.72

mberg

-0.71

ateral

-0.69

ittal

-0.68

itions

-0.68

Ground

-0.67

POSITIVE LOGITS

 Scion

0.70

 wolves

0.67

 vomit

0.67

 Assassins

0.66

 Forsaken

0.65

 bear

0.64

 incarcer

0.64

 overdoses

0.63

SAM

0.62

NRA

0.62

Activations Density 0.000%

special characters and formatting related to text adjustments

symbols or characters that signify special formatting

No Comments

No Known Activations