INDEX

Explanations

technical language related to operational protocols and assessments

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

ucci

-0.24

otton

-0.17

urse

-0.15

affen

-0.15

åĪĴ

-0.14

ura

-0.14

roys

-0.14

orse

-0.14

typeorm

-0.14

iferay

-0.14

POSITIVE LOGITS

 actual

0.28

actual

0.24

Actual

0.24

 Actual

0.23

_actual

0.21

 ìĭ¤ìłľ

0.20

å®ŀéĻħ

0.20

(actual

0.19

 concrete

0.19

 konkrÃ©t

0.19

Activations Density 0.210%

technical language related to operational protocols and assessments

No Comments

No Known Activations