INDEX

Explanations

reference to structured data formats or collections

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

raud

-0.16

conto

-0.16

itag

-0.15

elden

-0.14

eltas

-0.14

/autoload

-0.14

yonel

-0.14

ptions

-0.14

VERTISEMENT

-0.14

cdecl

-0.14

POSITIVE LOGITS

 throw

0.23

throw

0.20

 toss

0.18

 Throw

0.18

 thrown

0.18

 tossed

0.17

Throw

0.17

 tossing

0.16

pulse

0.16

	throw

0.14

Activations Density 0.118%

reference to structured data formats or collections

No Comments

No Known Activations