INDEX

Explanations

references to historical or academic topics and publications

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

zas

-0.17

ONS

-0.16

Ã¹a

-0.16

.animations

-0.16

ÑĥÑĢÑģ

-0.15

lisi

-0.15

batis

-0.15

conds

-0.14

ÑģÑĥÑĤ

-0.14

auge

-0.14

POSITIVE LOGITS

ota

0.15

.override

0.14

 Williamson

0.14

 Burger

0.14

 shemale

0.14

ÏĢÎµ

0.13

 Martinez

0.13

BOVE

0.13

::/

0.13

Activations Density 0.220%

references to historical or academic topics and publications

No Comments

No Known Activations