INDEX

Explanations

punctuation and formatting elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

egl

-0.17

olley

-0.16

 Trouble

-0.15

 Rencontre

-0.15

iland

-0.15

PORT

-0.14

ÙĤØ©

-0.14

ittel

-0.14

and

-0.14

za

-0.14

POSITIVE LOGITS

.createClass

0.15

 ÑģÐ²Ð¸

0.15

omit

0.15

ãĢĤæľ¬

0.14

 Mour

0.14

ela

0.14

wil

0.14

acco

0.14

ible

0.13

familia

0.13

Activations Density 0.041%

punctuation and formatting elements

No Comments

No Known Activations