INDEX

Explanations

occurrences of the word "last."

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

/Linux

-0.16

posite

-0.15

abus

-0.15

 escorte

-0.15

vat

-0.14

andin

-0.14

Ð¸Ð¸

-0.14

cken

-0.14

ANO

-0.14

QUEST

-0.14

POSITIVE LOGITS

-minute

0.42

est

0.35

minute

0.33

.fm

0.32

ingly

0.31

 resort

0.27

ly

0.27

IndexOf

0.26

 Resort

0.24

-second

0.24

Activations Density 0.061%

occurrences of the word "last."

No Comments

No Known Activations