INDEX

Explanations

references to Israel and its various contexts or attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

lug

-0.18

ulk

-0.16

agr

-0.16

resa

-0.16

au

-0.15

Ã§a

-0.15

ycz

-0.15

ucer

-0.15

lify

-0.14

piler

-0.14

POSITIVE LOGITS

 Defense

0.21

ites

0.20

ite

0.20

/P

0.20

rael

0.18

 lobby

0.18

lobby

0.17

ITE

0.17

led

0.17

 Bonds

0.17

Activations Density 0.017%

references to Israel and its various contexts or attributes

No Comments

No Known Activations