INDEX

Explanations

mathematical expressions and equations

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

oldur

-0.16

ģm

-0.16

Ã¡zi

-0.15

.MockMvc

-0.15

ÅĻeh

-0.15

xon

-0.15

elsey

-0.15

blank

-0.15

uzey

-0.15

abox

-0.14

POSITIVE LOGITS

 Struct

0.17

 Gross

0.16

ismus

0.15

 wings

0.15

interrupt

0.14

akin

0.14

-var

0.14

Activations Density 0.129%

mathematical expressions and equations

No Comments

No Known Activations