INDEX

Explanations

references to large families and their dynamics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

unte

-0.13

adesh

-0.13

gth

-0.13

isoft

-0.13

aphore

-0.12

 ë§ĲìĿ´

-0.12

 masturb

-0.12

aty

-0.12

eland

-0.12

odash

-0.12

POSITIVE LOGITS

 family

0.84

family

0.69

 families

0.69

 famil

0.67

fam

0.65

 Family

0.65

-family

0.64

 FAMILY

0.63

 familia

0.60

.family

0.59

Activations Density 0.147%

references to large families and their dynamics

No Comments

No Known Activations