INDEX

Explanations

time-related phrases and specific mentions of hours

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Late

-0.24

 late

-0.23

late

-0.21

 Nights

-0.20

Late

-0.20

 evenings

-0.19

 nights

-0.19

 overnight

-0.18

 Weeks

-0.18

 early

-0.17

POSITIVE LOGITS

 sharp

0.27

EST

0.25

 local

0.23

EST

0.22

ish

0.22

cst

0.21

 Sharp

0.21

ET

0.21

PST

0.20

ester

0.20

Activations Density 0.057%

time-related phrases and specific mentions of hours

No Comments

No Known Activations