INDEX

Explanations

terms related to weight and weight capacity

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

Ïį

-0.15

tid

-0.15

avery

-0.15

iq

-0.15

 timeouts

-0.14

è¡£

-0.14

 Duplicate

-0.14

ites

-0.14

idual

-0.14

umph

-0.14

POSITIVE LOGITS

 weight

0.19

weight

0.19

-weight

0.18

 Vest

0.17

weights

0.17

(weight

0.17

 weights

0.17

 heavier

0.16

 masses

0.15

Weight

0.15

Activations Density 0.076%

terms related to weight and weight capacity

No Comments

No Known Activations