INDEX

Explanations

references to specific events and actions related to rescue or emergency situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

åľ¨çº¿è§Ĩé¢ĳ

-0.16

ogui

-0.15

cona

-0.15

IRR

-0.15

errupt

-0.15

ueil

-0.15

ienes

-0.14

warts

-0.14

theid

-0.14

_Null

-0.14

POSITIVE LOGITS

ussen

0.14

ehler

0.14

due

0.14

à¸ģà¸£à¸ĵ

0.14

uyo

0.14

fal

0.14

Due

0.14

-collapse

0.14

 loss

0.14

innen

0.13

Activations Density 0.336%

references to specific events and actions related to rescue or emergency situations

No Comments

No Known Activations