INDEX

Explanations

descriptions of entertainers and their performance styles

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

ores

-0.17

-0.15

idal

-0.15

è¼

-0.14

CESS

-0.14

 discipline

-0.14

 ØºØ±

-0.14

RL

-0.13

nar

-0.13

uD

-0.13

POSITIVE LOGITS

 card

0.25

 spectator

0.22

 magic

0.22

/card

0.21

 cards

0.21

 spectators

0.20

 routines

0.20

 vanish

0.20

sle

0.20

 magician

0.20

Activations Density 0.019%

descriptions of entertainers and their performance styles

No Comments

No Known Activations