INDEX

Explanations

symbols or tokens from non-standard characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

uke

-0.15

Unicode

-0.14

ATTRIBUTE

-0.13

 ACLU

-0.13

Ĥæķ°

-0.13

 íĻľ

-0.12

church

-0.12

Advertis

-0.12

unc

-0.12

 interrog

-0.12

POSITIVE LOGITS

 Presidential

0.21

 Meeting

0.20

 transcript

0.18

 Transcript

0.18

ilateral

0.18

 Telephone

0.17

 Cabinet

0.17

Telephone

0.17

 meeting

0.17

 Meetings

0.17

Activations Density 0.005%

symbols or tokens from non-standard characters

No Comments

No Known Activations