INDEX

Explanations

references to self-related concepts and attributes

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

uez

-0.16

kaf

-0.14

aben

-0.14

ulia

-0.14

/framework

-0.14

ÑģÑĭÐ»

-0.14

pei

-0.13

ãĥ£

-0.13

 extent

-0.13

Dw

-0.13

POSITIVE LOGITS

/self

0.26

esteem

0.19

 self

0.19

Ã¤nd

0.19

 Self

0.18

sam

0.17

ridge

0.16

(Self

0.16

Î¯Î´Î±

0.16

eren

0.16

Activations Density 0.040%

references to self-related concepts and attributes

No Comments

No Known Activations