INDEX

Explanations

proper nouns related to specific entities like "Rogue," "Jude," "Derrick," and "ginger."

oai_token-act-pair · gpt-3.5-turbo

names and identifiers of people, places, or organizations

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

abilities

-0.81

atical

-0.78

alities

-0.78

ingham

-0.77

abul

-0.77

atically

-0.76

awan

-0.74

ada

-0.74

alysed

-0.74

aii

-0.73

POSITIVE LOGITS

 Ryder

0.76

 Baptist

0.72

 Ferdinand

0.69

Mars

0.67

 Lowe

0.66

 Trooper

0.65

bread

0.64

Dob

0.64

 Grimes

0.63

 Relations

0.63

Activations Density 0.113%

proper nouns related to specific entities like "Rogue," "Jude," "Derrick," and "ginger."

names and identifiers of people, places, or organizations

No Comments

No Known Activations