INDEX

Explanations

phrases that indicate transformation or conversion processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

amac

-0.15

beat

-0.15

nant

-0.14

cott

-0.14

unker

-0.14

pter

-0.14

dy

-0.14

celand

-0.13

quo

-0.13

boat

-0.13

POSITIVE LOGITS

0.16

/from

0.15

indr

0.15

Ã²

0.14

zÅĳ

0.14

áº½

0.13

olest

0.13

erializer

0.13

ovan

0.13

pdev

0.13

Activations Density 0.066%

phrases that indicate transformation or conversion processes

No Comments

No Known Activations