INDEX

Explanations

references to software releases and their characteristics

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

urd

-0.16

èĥŀ

-0.14

illard

-0.14

acements

-0.14

{text

-0.14

instein

-0.14

ibri

-0.14

zÄħ

-0.14

è¾ĳ

-0.13

.lu

-0.13

POSITIVE LOGITS

 Stable

0.39

 stable

0.38

stable

0.35

 unstable

0.33

 stabilization

0.33

 stability

0.31

 stabilized

0.31

ç¨³

0.30

RC

0.30

 beta

0.29

Activations Density 0.079%

references to software releases and their characteristics

No Comments

No Known Activations