INDEX

Explanations

references to race-related violence and exploitation

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

-0.14

irts

-0.13

crets

-0.13

nst

-0.13

ennon

-0.12

uilder

-0.12

ublished

-0.12

å°ĳå¥³

-0.12

 rains

-0.12

raj

-0.12

POSITIVE LOGITS

 expend

0.35

 disposable

0.32

 fodder

0.30

 chatt

0.30

 prey

0.28

 pawn

0.27

 cannon

0.27

 targets

0.26

 collateral

0.26

 Disposable

0.25

Activations Density 0.275%

references to race-related violence and exploitation

No Comments

No Known Activations