INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their qualifications or 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

FINE

-0.67

Weak

-0.66

 Meaning

-0.65

ÂłÂłÂłÂł

-0.65

INA

-0.65

 Saying

-0.64

 Opening

-0.64

--------------------------------------------------------

-0.64

++++++++

-0.63

Tank

-0.62

POSITIVE LOGITS

 specializes

1.57

 oversees

1.41

 teaches

1.38

 directs

1.37

 oversaw

1.37

 resear

1.25

 owns

1.21

 advises

1.21

 manages

1.19

 represents

1.16

Activations Density 0.104%

names of individuals

references to individuals and their qualifications or roles

No Comments

No Known Activations