INDEX

Explanations

terms related to quantities or groups of people

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

itos

-0.17

iq

-0.16

vais

-0.15

Ð½Ð¾ÑģÐ¸

-0.14

utdown

-0.14

ros

-0.14

 GlobalKey

-0.14

iams

-0.13

uthor

-0.13

addle

-0.13

POSITIVE LOGITS

 ones

0.16

 others

0.16

nul

0.16

ones

0.15

igaret

0.14

idle

0.14

rente

0.14

IFIC

0.13

 Ø§ÙĦØ®Ø§ØµØ©

0.13

 them

0.13

Activations Density 0.159%

terms related to quantities or groups of people

No Comments

No Known Activations