INDEX

Explanations

terms related to symmetry and asymmetry

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ering

-0.15

 Benton

-0.14

utton

-0.14

ÐºÑĥÐ»

-0.14

 ÑģÑħÐ¾Ð´

-0.13

ame

-0.13

ue

-0.13

Kid

-0.13

oga

-0.13

orado

-0.13

POSITIVE LOGITS

kaar

0.17

fan

0.17

edi

0.16

panies

0.16

posium

0.15

errupted

0.15

ãĤ¿ãĥ¼

0.15

ÑĩÐ°Ð¹

0.15

0.14

-minded

0.14

Activations Density 0.015%

terms related to symmetry and asymmetry

No Known Activations