INDEX

Explanations

symbols and punctuation used in lists or separations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

uple

-0.15

lad

-0.14

monds

-0.14

mented

-0.14

Fog

-0.14

aus

-0.13

tridges

-0.13

ToPoint

-0.13

ohon

-0.13

 æ²

-0.13

POSITIVE LOGITS

æķ¬

0.15

 Summon

0.15

Î³Î±

0.15

icers

0.15

EOS

0.15

 Trick

0.14

rÃ³d

0.14

 Heaven

0.14

egov

0.14

 Melee

0.14

Activations Density 0.038%

symbols and punctuation used in lists or separations

No Comments

No Known Activations