INDEX

Explanations

specific names or terms related to places, titles, or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 called

-0.21

called

-0.16

 Called

-0.15

cdn

-0.15

-called

-0.14

loquent

-0.14

 titled

-0.14

 nÃło

-0.14

Called

-0.14

à¸Ķà¸£

-0.14

POSITIVE LOGITS

 Ø§Ø®ØªØµ

0.24

 ÙĨØ³Ø¨Ø©

0.19

ãģ¨ãĤĤ

0.19

or

0.19

ÂĿ

0.19

 referring

0.18

 pronounced

0.17

 refer

0.16

 depending

0.16

 atau

0.15

Activations Density 0.182%

specific names or terms related to places, titles, or entities

No Comments

No Known Activations