INDEX

Explanations

the word "to" in various contexts related to intentions or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

ãģ¤ãģĳ

-0.20

è¦ĭ

-0.18

ãģ¤

-0.18

ToAdd

-0.16

nels

-0.16

éĸĭ

-0.15

@brief

-0.15

CastException

-0.15

gnore

-0.14

ĵåĲį

-0.14

POSITIVE LOGITS

/from

0.42

gether

0.32

asts

0.28

plevel

0.28

be

0.26

asting

0.25

-date

0.23

ying

0.22

pline

0.22

ady

0.22

Activations Density 0.784%

the word "to" in various contexts related to intentions or actions

No Comments

No Known Activations