INDEX

Explanations

various aspects of software evaluation and performance issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 Incontri

-0.13

 jadx

-0.13

 Continue

-0.12

(íģ¬ê¸°

-0.12

 kariy

-0.12

 rumpe

-0.11

 Angiospermae

-0.11

 ëĭ¤ìļ´ë°Ľê¸°

-0.11

 (“

-0.11

/>.↵↵

-0.11

POSITIVE LOGITS

åĪļæīį

0.15

 moden

0.13

.Footer

0.13

oriously

0.12

 devout

0.12

 aforementioned

0.12

 fraught

0.12

 yukarÄ±

0.11

.Ticks

0.11

 meticulous

0.11

Activations Density 0.048%

various aspects of software evaluation and performance issues

No Comments

No Known Activations