INDEX

Explanations

specific patterns or features in designs

oai_token-act-pair · gpt-3.5-turbo

references to various types of designs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Arist

-0.72

la

-0.67

 Elias

-0.66

 ILCS

-0.65

essional

-0.65

neau

-0.62

luc

-0.61

 charism

-0.60

 scient

-0.59

SPONSORED

-0.59

POSITIVE LOGITS

 designs

0.99

hops

0.91

paces

0.86

 layouts

0.86

heet

0.83

plates

0.82

isine

0.81

peed

0.78

pace

0.77

ators

0.76

Activations Density 0.020%

specific patterns or features in designs

references to various types of designs

No Comments

No Known Activations