INDEX

Explanations

references to torture and interrogation practices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.17.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

Î¹Î¿ÏĤ

-0.15

ê²°

-0.15

ByExample

-0.14

infeld

-0.14

alnÄ±z

-0.14

AspNet

-0.14

ÏĢÎ¿Ïħ

-0.14

apeutics

-0.14

_VIRTUAL

-0.14

nÃŃ

-0.14

POSITIVE LOGITS

 torture

0.42

 Tort

0.39

 tort

0.36

 TORT

0.34

 interrog

0.30

 tortured

0.28

 interrogation

0.27

-question

0.18

 questioning

0.18

 QUESTION

0.17

Activations Density 0.064%

references to torture and interrogation practices

No Comments

No Known Activations