INDEX

Explanations

critical keywords and terms expressing specific ideas or actions in a context

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

APT

-0.17

-0.16

hower

-0.15

ãĥ³ãĥģ

-0.15

Ð¾Ð»Ñİ

-0.15

apt

-0.14

hall

-0.14

ewe

-0.14

POSITIVE LOGITS

akra

0.17

reeNode

0.16

oose

0.15

Destructor

0.14

ycop

0.14

akis

0.14

esian

0.14

 HÃ²a

0.14

assis

0.14

Ð¶Ðµ

0.14

Activations Density 0.033%

critical keywords and terms expressing specific ideas or actions in a context

No Comments

No Known Activations