INDEX

Explanations

proper names and references related to historical or scholarly figures

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

arez

-0.15

 Hicks

-0.14

_bp

-0.14

 Suites

-0.14

ientes

-0.14

 mach

-0.14

uco

-0.14

.Accessible

-0.14

.AddComponent

-0.14

athers

-0.14

POSITIVE LOGITS

 Prem

0.15

 Extr

0.14

 Feather

0.14

extr

0.14

 Returned

0.13

apis

0.13

cheid

0.13

/cs

0.13

psc

0.13

edit

0.13

Activations Density 0.074%

proper names and references related to historical or scholarly figures

No Comments

No Known Activations