INDEX

Explanations

references to rewards or rewards-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to rewards or incentives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

abad

-0.77

abases

-0.75

 Ange

-0.72

Osw

-0.69

lander

-0.69

obiles

-0.68

enium

-0.67

 Sloven

-0.66

ahime

-0.64

inx

-0.63

POSITIVE LOGITS

 reward

1.08

 rewards

1.01

 rewarded

0.97

 rewarding

0.79

Reward

0.78

 recipients

0.75

 tiers

0.74

 Reward

0.73

 incentive

0.73

 fulfillment

0.72

Activations Density 0.020%

references to rewards or rewards-related terms

terms related to rewards or incentives

No Comments

No Known Activations