INDEX

Explanations

instances where control or empowerment is emphasized in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

uario

-0.17

iasi

-0.16

ings

-0.15

ë¡ľëĵľ

-0.14

 ä»£

-0.14

ãĥ¬ãĥĥãĥĪ

-0.14

 addCriterion

-0.14

_PRIORITY

-0.14

 Shock

-0.13

ollapse

-0.13

POSITIVE LOGITS

/control

0.26

 control

0.24

(control

0.22

 destiny

0.21

.Control

0.20

-control

0.20

 CONTROL

0.20

 Control

0.20

control

0.19

 controlled

0.19

Activations Density 0.059%

instances where control or empowerment is emphasized in various contexts

No Comments

No Known Activations