INDEX

Explanations

references to confidential or sensitive information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

igu

-0.14

issan

-0.14

 excess

-0.14

ÑĸÐ¾Ð½

-0.13

anto

-0.13

ennis

-0.13

 dignity

-0.12

 tisk

-0.12

ç·Ĵ

-0.12

ALES

-0.12

POSITIVE LOGITS

 leaked

0.40

 leak

0.37

 obtained

0.35

 Leak

0.34

 leaks

0.33

 documents

0.28

æ³

0.28

 released

0.27

 leaking

0.26

Obt

0.26

Activations Density 0.214%

references to confidential or sensitive information

No Comments

No Known Activations