INDEX

Explanations

references to breaks and hiatuses 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 @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 Rencontres

-0.14

 overlooked

-0.14

URRE

-0.14

 outf

-0.13

erton

-0.13

aki

-0.13

erea

-0.13

Utf

-0.13

Undo

-0.13

 Undo

-0.13

POSITIVE LOGITS

 hiatus

0.46

 break

0.43

 pause

0.43

ä¼ĳ

0.37

 breaks

0.37

 Pause

0.37

pause

0.35

 pauses

0.34

Pause

0.33

.pause

0.31

Activations Density 0.219%

references to breaks and hiatuses in various contexts

No Comments

No Known Activations