INDEX

Explanations

instances of the word "ret" and its variations within a context indicating retrieval or referencing

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

erland

-0.17

oles

-0.16

pare

-0.15

unting

-0.15

heit

-0.15

patial

-0.15

serter

-0.15

raise

-0.15

ieder

-0.14

#Region

-0.14

POSITIVE LOGITS

Ret

0.19

angular

0.19

ted

0.18

(ret

0.18

.Ret

0.17

ters

0.17

tings

0.16

-ret

0.16

 Jordan

0.15

ention

0.15

Activations Density 0.032%

instances of the word "ret" and its variations within a context indicating retrieval or referencing

No Comments

No Known Activations