INDEX

Explanations

references to towers and their architectural features

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

Ð¾ÑģÑĥÐ´

-0.16

 Valle

-0.16

oad

-0.15

Ãºmer

-0.14

alach

-0.14

lfw

-0.14

ollar

-0.14

ortal

-0.14

Tek

-0.14

Ð¼Ð¸ÑĤ

-0.13

POSITIVE LOGITS

fault

0.17

hart

0.15

Ø§Ø°

0.14

MethodInfo

0.14

ÙĨØ´

0.14

à¸£à¸°

0.14

isha

0.14

 taller

0.14

 tall

0.13

 Hari

0.13

Activations Density 0.023%

references to towers and their architectural features

No Comments

No Known Activations