INDEX

Explanations

references to risky actions or decisions

oai_token-act-pair · gpt-3.5-turbo

phrases related to gambling and risky decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

NS

-0.67

 stout

-0.65

 indexed

-0.64

Autom

-0.64

 citations

-0.63

 stub

-0.62

 metro

-0.61

 faire

-0.61

 leveled

-0.60

 flat

-0.60

POSITIVE LOGITS

 gamble

3.01

 Catch

1.67

 Gamble

1.59

 gamb

1.44

 dilemma

1.24

hang

1.16

exist

1.09

 risky

1.07

 balancing

0.98

 calculus

0.95

Activations Density 0.057%

references to risky actions or decisions

phrases related to gambling and risky decisions

No Comments

No Known Activations