INDEX

Explanations

phrases indicating capability and assistance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

anyak

-0.17

okud

-0.15

.sponge

-0.14

 lÃłnh

-0.14

ropolitan

-0.14

dict

-0.13

åķĬåķĬ

-0.13

aza

-0.13

ÅĽmy

-0.13

reak

-0.13

POSITIVE LOGITS

arah

0.15

jure

0.15

apus

0.15

adel

0.14

nap

0.14

Mek

0.14

Jar

0.14

 diseÃ±ador

0.13

_userdata

0.13

Activations Density 0.078%

phrases indicating capability and assistance

No Comments

No Known Activations