INDEX

Explanations

references to websites or applications

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

yro

-0.15

inka

-0.15

zbo

-0.15

 sands

-0.14

anford

-0.14

 Baldwin

-0.14

agal

-0.14

plural

-0.14

ido

-0.13

benchmark

-0.13

POSITIVE LOGITS

Map

0.20

nek

0.18

Map

0.17

-map

0.17

Vak

0.17

map

0.17

MAP

0.16

map

0.16

ody

0.15

_Map

0.15

Activations Density 0.012%

references to websites or applications

No Comments

No Known Activations