INDEX

Explanations

references to communication and group discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

atab

-0.16

telegram

-0.15

atoi

-0.14

ONUS

-0.14

ANEL

-0.14

------------------------------------------------------------------------------------------------

-0.14

ÑģÐ»Ð¾Ð²

-0.14

orns

-0.14

é½

-0.13

hus

-0.13

POSITIVE LOGITS

 Subject

0.25

 unsubscribe

0.22

unsubscribe

0.22

Subject

0.21

 Digest

0.21

 digest

0.20

 Archives

0.20

 bounce

0.20

ê´ĳê³ł

0.19

 Sent

0.19

Activations Density 0.100%

references to communication and group discussions

No Comments

No Known Activations