INDEX

Explanations

phrases emphasizing the word "quite" followed by articles like "a" or "an"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ÐµÑĢÑĪ

-0.15

chg

-0.15

 somehow

-0.14

hist

-0.14

ialis

-0.14

isis

-0.14

imon

-0.14

qv

-0.14

ivre

-0.13

rep

-0.13

POSITIVE LOGITS

-ÑĤÐ°ÐºÐ¸

0.20

acket

0.19

oker

0.16

ylum

0.16

PAIR

0.15

ahir

0.15

vron

0.14

ournaments

0.14

 quot

0.14

Ð²Ð°Ð¹

0.14

Activations Density 0.021%

phrases emphasizing the word "quite" followed by articles like "a" or "an"

No Comments

No Known Activations