INDEX

Explanations

names of people and characters, particularly in a context related to performance or competition

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

achten

-0.16

uman

-0.16

etable

-0.15

usu

-0.14

itech

-0.14

.getID

-0.14

lassen

-0.14

Î¼ÏĢÎ¿

-0.14

berman

-0.14

igy

-0.13

POSITIVE LOGITS

eer

0.19

ius

0.17

ously

0.15

asca

0.15

cki

0.14

Ù

0.14

ment

0.14

oms

0.14

initely

0.14

vers

0.14

Activations Density 0.127%

names of people and characters, particularly in a context related to performance or competition

No Comments

No Known Activations

names of people and characters, particularly in a context related to performance or competition

No Comments

No Known Activations