INDEX

Explanations

structured data or log entries, often related to software versions or timestamps

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 Carrier

-0.18

Ð°Ð½ÑĮ

-0.16

.CONTENT

-0.14

Carrier

-0.14

ym

-0.14

thing

-0.14

bif

-0.13

leaf

-0.13

uestion

-0.13

kup

-0.13

POSITIVE LOGITS

ogi

0.17

 deser

0.17

 Chandler

0.16

.ActionListener

0.15

Appending

0.15

ersist

0.15

.VK

0.14

Uvs

0.14

ecz

0.14

/op

0.14

Activations Density 0.005%

structured data or log entries, often related to software versions or timestamps

No Comments

No Known Activations