INDEX

Explanations

statements made by spokespersons in various situations

oai_token-act-pair · gpt-3.5-turbo

mentions of spokespeople and their roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ised

-0.73

[];

-0.72

 Pell

-0.67

OUS

-0.67

 Haram

-0.66

 Bagg

-0.64

ITIES

-0.62

ising

-0.60

 Hubble

-0.60

 Apocalypse

-0.60

POSITIVE LOGITS

men

2.64

man

2.61

mans

2.11

woman

2.08

women

1.99

manship

1.96

person

1.85

people

1.83

mens

1.73

MAN

1.59

Activations Density 0.232%

statements made by spokespersons in various situations

mentions of spokespeople and their roles

No Comments

No Known Activations