INDEX

Explanations

references to dishonor and the consequences of betrayal

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 gave

-0.33

 threw

-0.31

 drew

-0.30

 wrote

-0.29

 grew

-0.28

 blew

-0.28

saw

-0.28

 took

-0.28

 broke

-0.26

took

-0.25

POSITIVE LOGITS

 taken

0.40

 gone

0.40

 seen

0.38

 spoken

0.37

 gotten

0.36

 flown

0.36

 Seen

0.35

 Taken

0.35

 eaten

0.34

idden

0.33

Activations Density 0.117%

references to dishonor and the consequences of betrayal

No Comments

No Known Activations