INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

the presence of specific names, particularly related to well-known figures

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-0.77

 Dartmouth

-0.74

oster

-0.72

chie

-0.70

 refere

-0.67

 Islanders

-0.67

inances

-0.65

ingham

-0.64

 Kling

-0.62

EEK

-0.62

POSITIVE LOGITS

 Sonia

1.20

 Gandhi

0.92

eus

0.88

uria

0.84

inia

0.83

":""},{"

0.82

ata

0.78

OTUS

0.75

otom

0.73

 Tome

0.72

Activations Density 0.012%

names of individuals

the presence of specific names, particularly related to well-known figures

No Comments

No Known Activations