INDEX

Explanations

programming and networking terminology

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

onz

-0.17

ritz

-0.16

itsu

-0.15

 Riders

-0.15

chrift

-0.15

-0.14

lesia

-0.14

 Damen

-0.14

ATS

-0.14

ynamic

-0.14

POSITIVE LOGITS

 pneum

0.21

Jo

0.20

 driv

0.19

 intake

0.19

 commanded

0.18

jo

0.18

tal

0.18

 serv

0.17

enc

0.17

_PID

0.17

Activations Density 0.013%

programming and networking terminology

No Comments

No Known Activations