INDEX

Explanations

references to various "rules" or "rule" systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

omes

-0.18

ienne

-0.17

ness

-0.17

imals

-0.16

Ú¯Ø§Ùĩ

-0.16

Å¼e

-0.16

sein

-0.16

/sn

-0.15

rix

-0.15

Ð¸Ð¸

-0.15

POSITIVE LOGITS

book

0.30

 thumb

0.27

books

0.23

making

0.22

 Thumb

0.20

ãģ¨ãģĵãĤį

0.20

ets

0.20

thumb

0.19

enstein

0.18

Thumb

0.18

Activations Density 0.026%

references to various "rules" or "rule" systems

No Comments

No Known Activations