INDEX

Explanations

design-related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

agal

-0.17

arer

-0.16

pire

-0.16

 pitch

-0.15

 Manga

-0.15

ç¦ıåĪ©

-0.15

 Pitch

-0.15

sti

-0.15

izons

-0.14

erland

-0.14

POSITIVE LOGITS

uran

0.19

Pow

0.18

pow

0.17

Pow

0.17

pow

0.17

 Bron

0.16

asset

0.16

 Warm

0.15

gm

0.15

Warm

0.15

Activations Density 0.007%

design-related terminology

No Comments

No Known Activations