INDEX

Explanations

proper nouns or titles, particularly related to political figures or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

verts

-0.16

ihn

-0.15

ound

-0.15

Ð°Ð¼Ð¿

-0.14

Pam

-0.14

ogie

-0.14

keh

-0.14

oundary

-0.14

lix

-0.14

ete

-0.14

POSITIVE LOGITS

Ã¤s

0.34

Ã¤

0.26

Ã¤sent

0.25

inz

0.25

unks

0.23

unk

0.23

akt

0.22

isen

0.21

aes

0.21

Ã¤h

0.20

Activations Density 0.006%

proper nouns or titles, particularly related to political figures or entities

No Comments

No Known Activations