INDEX

Explanations

instances of humor and teasing in conversation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

ania

-0.15

');?>

-0.14

utomation

-0.14

 Cortex

-0.14

ixo

-0.14

 Morav

-0.13

 Plantae

-0.13

rant

-0.13

 Alexis

-0.13

è¨³

-0.12

POSITIVE LOGITS

 Baum

0.15

afone

0.15

 ØºØ±

0.15

ingleton

0.14

azen

0.14

ØºØ±

0.14

arkin

0.14

 shorts

0.14

 lightning

0.14

foon

0.14

Activations Density 0.182%

instances of humor and teasing in conversation

No Comments

No Known Activations