INDEX

Explanations

reference markers or components typically used in programming or code snippets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

rait

-0.16

Wax

-0.15

ake

-0.14

 conservatives

-0.14

uff

-0.14

umuz

-0.14

__

-0.14

re

-0.14

ions

-0.14

afort

-0.14

POSITIVE LOGITS

æµ©

0.18

ramework

0.17

ipay

0.16

MethodInfo

0.15

rido

0.15

iglia

0.15

XHR

0.14

Slf

0.14

typings

0.14

kie

0.14

Activations Density 0.001%

reference markers or components typically used in programming or code snippets

No Comments

No Known Activations