INDEX

Explanations

proper nouns related to legal or political figures

oai_token-act-pair · gpt-3.5-turbo

names and references related to specific individuals, particularly those involved in legal or political issues

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

 Farming

-0.82

 Generations

-0.78

 Glass

-0.74

\\\\\\\\

-0.73

 Curious

-0.71

 Terr

-0.70

BuyableInstoreAndOnline

-0.70

atility

-0.70

 Percy

-0.70

edIn

-0.69

POSITIVE LOGITS

uminati

0.98

ivot

0.91

uxe

0.83

rogens

0.82

cial

0.80

endar

0.79

zona

0.79

rogen

0.78

alyst

0.75

eval

0.72

Activations Density 0.020%

proper nouns related to legal or political figures

names and references related to specific individuals, particularly those involved in legal or political issues

No Comments

No Known Activations