INDEX

Explanations

references to volumes or issues of a publication, likely in a periodical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

elong

-0.17

_DEAD

-0.14

insk

-0.14

 Odyssey

-0.14

 tier

-0.14

ometers

-0.14

ensa

-0.14

Î¸Î·

-0.13

ippi

-0.13

UNT

-0.13

POSITIVE LOGITS

ÅĽnie

0.19

Prog

0.16

conde

0.15

ureka

0.15

 Ð´Ð¾ÑĢÐ¾Ð³

0.14

Him

0.14

 odbor

0.14

esign

0.14

ockey

0.14

 Ø§ÙĦØ§Ø¹

0.14

Activations Density 0.005%

references to volumes or issues of a publication, likely in a periodical context

No Comments

No Known Activations