INDEX

Explanations

references to competitive events or awards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

udic

-0.15

 meas

-0.14

å¿ħ

-0.14

aval

-0.14

Aux

-0.14

æ¦Ĥ

-0.14

ÐµÐºÑģ

-0.13

mÃŃn

-0.13

rost

-0.13

Binding

-0.13

POSITIVE LOGITS

ofday

0.18

mur

0.15

UnderTest

0.15

ÐºÐ¾ÑĢÐ¸ÑģÑĤ

0.15

eral

0.15

Ä±lÄ±

0.14

lama

0.14

ersiz

0.14

_NAMESPACE

0.14

ulumi

0.14

Activations Density 0.156%

references to competitive events or awards

No Comments

No Known Activations