INDEX

Explanations

references to cybersecurity and government-related programs or initiatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ÑĢÐ°Ð±

-0.19

ç¿

-0.15

adelphia

-0.15

Æ°ng

-0.14

Ð»Ð°ÑĤ

-0.14

å¤©åłĤ

-0.14

ama

-0.14

afil

-0.14

oÅ¾

-0.14

POSITIVE LOGITS

Oak

0.39

DOE

0.36

Oak

0.32

 Liver

0.30

DO

0.29

 Lawrence

0.27

Arg

0.25

Liver

0.25

 Brook

0.24

oak

0.24

Activations Density 0.050%

references to cybersecurity and government-related programs or initiatives

No Comments

No Known Activations