INDEX

Explanations

names of historical or notable individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

Fat

-0.17

kola

-0.16

lington

-0.16

ATO

-0.15

egin

-0.14

fat

-0.14

angelo

-0.14

neh

-0.14

ato

-0.14

elas

-0.14

POSITIVE LOGITS

_AUX

0.16

sonian

0.16

pg

0.15

craft

0.15

 voks

0.14

 Horny

0.14

çŃĴ

0.14

ãĥ¼ãĥ«

0.14

(iOS

0.13

riel

0.13

Activations Density 0.140%

names of historical or notable individuals

No Comments

No Known Activations