INDEX

Explanations

phrases where someone or something is identified as a specific category or label

oai_token-act-pair · gpt-3.5-turbo

phrases indicating identity or classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

amon

-0.75

rones

-0.73

reprene

-0.72

ongyang

-0.70

raq

-0.67

thumbnails

-0.65

iban

-0.64

iger

-0.63

ffer

-0.63

apons

-0.63

POSITIVE LOGITS

 belonging

1.14

 having

1.03

 follows

1.03

 needing

0.91

 being

0.90

pires

0.90

 such

0.86

 possessing

0.86

 either

0.82

 originating

0.81

Activations Density 0.113%

phrases where someone or something is identified as a specific category or label

phrases indicating identity or classification

No Comments

No Known Activations