INDEX

Explanations

references to death or lethal events

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

 bÃło

-0.18

 Trab

-0.15

 Disorder

-0.14

Ð¾Ð»ÐµÐ²Ð°

-0.14

HRESULT

-0.14

ÃŃme

-0.14

asurer

-0.14

lfw

-0.14

ÅĻes

-0.14

ancell

-0.13

POSITIVE LOGITS

die

0.65

 died

0.59

 dies

0.56

die

0.52

DIE

0.49

Die

0.48

_die

0.47

Die

0.46

 dying

0.45

 Dies

0.41

Activations Density 0.213%

references to death or lethal events

No Comments

No Known Activations