INDEX

Explanations

references to historical buildings and cultural sites

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

 Druid

-0.17

 Apost

-0.15

Pri

-0.15

 SavaÅŁ

-0.15

regor

-0.15

 Evangel

-0.14

 VÅ¡ech

-0.14

onden

-0.14

 Krishna

-0.14

ãĤ·ãĥ¼

-0.14

POSITIVE LOGITS

dj

0.22

 carpets

0.21

 camel

0.19

 genie

0.19

Nas

0.19

Bed

0.18

Saf

0.18

rug

0.18

 rugs

0.18

Activations Density 0.373%

references to historical buildings and cultural sites

No Comments

No Known Activations