INDEX

Explanations

words associated with entertainment or media content

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

InstanceId

-0.15

anship

-0.15

bedo

-0.15

 Warfare

-0.14

ãĥ¼ãĥ©

-0.14

armacy

-0.14

ypress

-0.14

 McCart

-0.14

raith

-0.14

POSITIVE LOGITS

lopen

0.16

.heroku

0.16

 Grande

0.15

iated

0.14

ADV

0.14

PD

0.14

jn

0.14

umbnails

0.14

PN

0.14

irs

0.14

Activations Density 0.000%

words associated with entertainment or media content

No Comments

No Known Activations