INDEX

Explanations

terminology related to rewards and incentives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ernote

-0.15

esting

-0.15

ault

-0.15

ellas

-0.14

UNCH

-0.14

ushing

-0.14

Dud

-0.14

SmartPointer

-0.14

 Pessoa

-0.14

ushed

-0.13

POSITIVE LOGITS

 reward

0.25

 rewards

0.23

 Rewards

0.20

 rewarded

0.20

reward

0.19

 Reward

0.18

 rewarding

0.18

plier

0.16

ple

0.15

.impl

0.15

Activations Density 0.104%

terminology related to rewards and incentives

No Comments

No Known Activations