INDEX

Explanations

phrases related to winning and performance in competitive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

çŃĴ

-0.17

anel

-0.16

atre

-0.16

ugo

-0.14

apol

-0.14

Ð°Ð»Ð¾Ð¼

-0.13

acles

-0.13

 Paso

-0.13

mlx

-0.13

ç½²

-0.13

POSITIVE LOGITS

 Wilkinson

0.14

OMP

0.14

ìĹħ

0.14

 Emmy

0.13

 Downing

0.13

Ð¾Ð»Ð½

0.13

nul

0.13

 down

0.13

 Kingdom

0.13

upp

0.13

Activations Density 0.084%

phrases related to winning and performance in competitive contexts

No Comments

No Known Activations