INDEX

Explanations

references to violence or harmful events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

bout

-0.16

UME

-0.15

agger

-0.14

 Spicer

-0.14

ãĥ¼ãĥį

-0.14

exo

-0.14

ugo

-0.14

ereo

-0.13

stoupil

-0.13

POSITIVE LOGITS

kdo

0.15

edd

0.15

Verdana

0.14

inous

0.14

igs

0.14

ysz

0.14

yre

0.14

 Abdullah

0.14

 Beauty

0.13

dcc

0.13

Activations Density 0.237%

references to violence or harmful events

No Comments

No Known Activations