INDEX

Explanations

phrases related to rewards and incentives

oai_token-act-pair · gpt-3.5-turbo

mentions of "rewards."

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

vor

-0.70

ni

-0.69

rano

-0.64

law

-0.64

vi

-0.63

Sy

-0.63

 Column

-0.62

dc

-0.61

FORM

-0.61

 Univ

-0.61

POSITIVE LOGITS

 rewards

3.84

 reward

2.96

 Rewards

2.57

 Reward

2.36

Reward

2.12

 prizes

2.03

 rewarded

1.92

 rewarding

1.82

 bonuses

1.70

 incentives

1.69

Activations Density 0.032%

phrases related to rewards and incentives

mentions of "rewards."

No Comments

No Known Activations