INDEX

Explanations

structured lists or items, especially in the context of advice or recommendations

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

bane

-0.17

ÑĥÐ´Ð°

-0.15

weit

-0.15

oux

-0.15

abar

-0.15

zia

-0.14

èĬ¯

-0.14

åĩĮ

-0.13

laden

-0.13

-scripts

-0.13

POSITIVE LOGITS

alink

0.16

è®®

0.16

ettle

0.15

suma

0.15

edd

0.14

<+

0.14

desk

0.14

aiser

0.14

lite

0.14

Ly

0.13

Activations Density 0.042%

structured lists or items, especially in the context of advice or recommendations

No Comments

No Known Activations