INDEX

Explanations

direct address to the reader, emphasizing suggestions and advice

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

-0.15

Abs

-0.14

isu

-0.14

Met

-0.14

estre

-0.13

zel

-0.13

 Ø±Ùħ

-0.13

 Ð¾ÐºÐ°Ð·

-0.13

 Happ

-0.13

Ã¼h

-0.13

POSITIVE LOGITS

Should

0.17

 should

0.16

(çģ«

0.16

should

0.16

Debe

0.16

 ought

0.16

_should

0.15

aks

0.15

ogra

0.15

iddet

0.14

Activations Density 0.053%

direct address to the reader, emphasizing suggestions and advice

No Comments

No Known Activations