INDEX

Explanations

cooking-related instructions and actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

ancel

-0.17

ynet

-0.16

unden

-0.14

exus

-0.14

 Various

-0.14

å¤ļ

-0.14

orian

-0.14

 multiple

-0.14

ence

-0.13

 Multiple

-0.13

POSITIVE LOGITS

 reserved

0.23

Reserved

0.21

 Reserved

0.21

Remaining

0.21

 remaining

0.21

reserved

0.21

remaining

0.19

 Remaining

0.19

 HALF

0.19

 drained

0.17

Activations Density 0.038%

cooking-related instructions and actions

No Comments

No Known Activations