INDEX

Explanations

imperative verbs and action-oriented phrases related to processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 pornÃ´s

-0.20

pone

-0.16

 Prostitutas

-0.16

hiba

-0.14

odash

-0.14

PageRoute

-0.13

 Millenn

-0.13

neas

-0.13

nye

-0.13

Ã¨ve

-0.13

POSITIVE LOGITS

ser

0.34

 fazer

0.31

ter

0.29

 criar

0.26

dar

0.26

 estar

0.26

 contar

0.25

 entrar

0.25

 realizar

0.24

 tomar

0.24

Activations Density 0.011%

imperative verbs and action-oriented phrases related to processes

No Comments

No Known Activations