INDEX

Explanations

references to specific films and their associated titles or characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

afka

-0.15

 Grupo

-0.14

ycin

-0.14

ople

-0.14

enter

-0.14

anca

-0.13

jer

-0.13

 vÃ½ro

-0.13

Ð´ÐµÐ¹ÑģÑĤÐ²

-0.13

rowable

-0.13

POSITIVE LOGITS

 Confidential

0.17

ãģĻãģĻ

0.15

amma

0.14

 Wars

0.14

ìī

0.14

 expend

0.14

 supporting

0.14

å®ħ

0.14

ropa

0.14

Activations Density 0.250%

references to specific films and their associated titles or characters

No Comments

No Known Activations