INDEX

Explanations

instances of editorial references or notes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ayers

-0.16

heimer

-0.16

áº¡ng

-0.15

PARAM

-0.15

ahy

-0.14

ever

-0.14

 param

-0.14

ÑĦÐµÑĢ

-0.14

oro

-0.14

erox

-0.14

POSITIVE LOGITS

ials

0.35

ial

0.33

IAL

0.28

/editor

0.27

ially

0.27

iale

0.24

(Editor

0.24

 editor

0.23

-editor

0.23

iales

0.23

Activations Density 0.019%

instances of editorial references or notes

No Comments

No Known Activations