INDEX

Explanations

terms associated with value and evaluation in various contexts

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

xcf

-0.16

ÙĨØ¯Ùĩ

-0.14

 abduction

-0.14

 Lair

-0.14

oscope

-0.14

istrovstvÃŃ

-0.13

fak

-0.13

 ConnectionState

-0.13

gross

-0.13

AVING

-0.13

POSITIVE LOGITS

/on

0.15

ÄĻÅ¼

0.14

Saf

0.13

ettle

0.13

eous

0.13

occo

0.13

kind

0.13

ÃŃo

0.13

ceptar

0.13

erra

0.13

Activations Density 0.368%

terms associated with value and evaluation in various contexts

No Comments

No Known Activations