INDEX

Explanations

references to educational or informational resources and leadership concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ras

-0.15

ascar

-0.15

arel

-0.14

æİĴ

-0.14

ä¿

-0.14

rott

-0.14

assium

-0.14

à¤Ĺà¤²

-0.14

 sizing

-0.13

abras

-0.13

POSITIVE LOGITS

DLC

0.20

 hoop

0.19

chief

0.18

LC

0.17

--[

0.16

ãĤªãĥ³

0.16

OPY

0.16

istrovstvÃŃ

0.15

iku

0.15

lc

0.15

Activations Density 0.031%

references to educational or informational resources and leadership concepts

No Comments

No Known Activations