INDEX

Explanations

terms related to communication or coordination roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

yles

-0.17

spread

-0.16

BUFF

-0.16

 shedding

-0.16

vrier

-0.15

til

-0.14

ukes

-0.14

empo

-0.14

 Weiner

-0.14

ãĥ³ãĤ¯

-0.14

POSITIVE LOGITS

ison

0.35

isons

0.35

ISON

0.29

aison

0.23

ision

0.22

ising

0.21

quat

0.19

ise

0.19

izon

0.18

SION

0.18

Activations Density 0.004%

terms related to communication or coordination roles

No Comments

No Known Activations