INDEX

Explanations

phrases related to geographic locations or political jurisdictions

oai_token-act-pair · gpt-3.5-turbo

terms related to geographic policies or actions at a broad scale

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

wright

-0.91

ion

-0.86

ioned

-0.82

rett

-0.74

module

-0.73

nell

-0.72

ions

-0.72

emis

-0.71

poons

-0.69

 Ital

-0.67

POSITIVE LOGITS

wide

0.99

 unemployment

0.80

 blackout

0.75

 distribution

0.75

 pandemonium

0.75

 condemnation

0.74

ensus

0.74

 acclaim

0.72

 outcry

0.71

 broadband

0.71

Activations Density 0.043%

phrases related to geographic locations or political jurisdictions

terms related to geographic policies or actions at a broad scale

No Comments

No Known Activations