INDEX

Explanations

terms related to layering or organization, particularly involving the word "deck."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

-thirds

-0.17

fter

-0.16

istic

-0.16

ÑģÐµÐ½

-0.15

olen

-0.15

ynch

-0.14

ous

-0.14

atic

-0.14

ogue

-0.14

arians

-0.14

POSITIVE LOGITS

NOWLED

0.22

nowledge

0.21

nowled

0.21

../../../

0.20

enzie

0.19

eeper

0.19

abee

0.18

tober

0.18

tails

0.18

al

0.18

Activations Density 0.160%

terms related to layering or organization, particularly involving the word "deck."

No Comments

No Known Activations