INDEX

Explanations

words related to self-referential concepts and actions

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

abwe

-0.16

ahoma

-0.16

iform

-0.15

akest

-0.14

geb

-0.14

argent

-0.14

 Jensen

-0.14

tie

-0.14

idia

-0.13

éĪ

-0.13

POSITIVE LOGITS

/self

0.41

 Self

0.33

Self

0.30

 self

0.29

self

0.28

(Self

0.28

 SELF

0.27

-self

0.26

SELF

0.25

=self

0.23

Activations Density 0.031%

words related to self-referential concepts and actions

No Comments

No Known Activations