INDEX

Explanations

variations of the word related to winning or achievement

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

avors

-0.16

ãĥ¬ãĥ³

-0.16

.Completed

-0.15

ottage

-0.15

ottes

-0.15

ä¼ı

-0.15

geh

-0.15

uito

-0.15

mine

-0.14

rawer

-0.14

POSITIVE LOGITS

alt

0.21

inn

0.20

inner

0.20

ei

0.20

oll

0.18

orden

0.18

Gew

0.18

ri

0.18

ona

0.17

altar

0.17

Activations Density 0.005%

variations of the word related to winning or achievement

No Comments

No Known Activations