INDEX

Explanations

references to time and duration in experiences or roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

uzzi

-0.18

stdcall

-0.16

ziej

-0.15

ÐµÑĩÐ½Ð¾

-0.15

ulumi

-0.15

lates

-0.14

typings

-0.14

ummer

-0.14

ordes

-0.14

romo

-0.14

POSITIVE LOGITS

 tenure

0.49

 stay

0.45

 time

0.35

 reign

0.33

stay

0.32

 stays

0.31

 stint

0.31

 Stay

0.28

Stay

0.27

 existence

0.25

Activations Density 0.101%

references to time and duration in experiences or roles

No Comments

No Known Activations