INDEX

Explanations

elements related to connections and agreements in discussions or analyses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Include

-0.20

 include

-0.20

include

-0.19

 cannot

-0.18

 Cannot

-0.17

 INCLUDE

-0.17

Include

-0.17

 contain

-0.17

 certainly

-0.17

.include

-0.17

POSITIVE LOGITS

 compares

0.33

 affect

0.33

 affects

0.33

 relates

0.33

 affected

0.32

 compare

0.31

 fits

0.30

 fares

0.30

 afect

0.30

fit

0.30

Activations Density 0.106%

elements related to connections and agreements in discussions or analyses

No Comments

No Known Activations