INDEX

Explanations

references to the phrase "Lord of the Rings."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

assi

-0.15

 Beste

-0.15

visor

-0.15

ãĥĥãĤ¯ãĤ¹

-0.15

 dele

-0.14

 Barb

-0.14

upt

-0.14

aft

-0.14

 Norse

-0.14

cab

-0.14

POSITIVE LOGITS

 rings

0.17

jedn

0.16

 ring

0.16

owell

0.16

gv

0.15

asje

0.15

Î´Î®

0.14

ekli

0.14

 Ring

0.14

imli

0.14

Activations Density 0.018%

references to the phrase "Lord of the Rings."

No Comments

No Known Activations