INDEX

Explanations

phrases that emphasize the superlative or highlight notable subjects or items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

 èı²å¾ĭå®¾

-0.15

aret

-0.15

<typeof

-0.15

_Impl

-0.15

.rec

-0.15

ÑĢÑĸÐ¹

-0.15

ONTAL

-0.14

 Ù¾ÛĮÚ©

-0.14

ayet

-0.14

»

-0.14

POSITIVE LOGITS

.cloud

0.16

sms

0.15

SM

0.14

strap

0.14

cap

0.14

ç²Ĺ

0.14

adow

0.14

 simple

0.14

est

0.14

ew

0.13

Activations Density 0.018%

phrases that emphasize the superlative or highlight notable subjects or items

No Comments

No Known Activations