INDEX

Explanations

numerical statistics or data representations

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

åŀ

-0.17

aliz

-0.16

 Colo

-0.15

chers

-0.15

ClassLoader

-0.14

uliar

-0.14

 equal

-0.14

ames

-0.13

POSITIVE LOGITS

-LAST

0.17

isArray

0.16

alach

0.15

Ø¶ÙĪ

0.15

ï»¿#

0.15

bat

0.14

cimal

0.14

ois

0.14

eger

0.14

TemplateName

0.14

Activations Density 0.016%

numerical statistics or data representations

No Comments

No Known Activations