INDEX

Explanations

information related to error messages and policy validation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

emez

-0.17

nze

-0.16

outu

-0.15

+:+

-0.15

oq

-0.15

esub

-0.15

ubern

-0.15

onom

-0.15

eldorf

-0.14

ourg

-0.14

POSITIVE LOGITS

/jav

0.17

0.16

ount

0.14

Dual

0.14

å®¶

0.14

Nam

0.14

0.13

upo

0.13

Activations Density 0.017%

information related to error messages and policy validation

No Comments

No Known Activations