INDEX

Explanations

words and phrases associated with violence and physical harm

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pty

-0.17

/boot

-0.15

ies

-0.15

aly

-0.15

eme

-0.15

ales

-0.14

ulet

-0.14

Ð¸Ð»Ñı

-0.14

ale

-0.14

ilir

-0.14

POSITIVE LOGITS

ĽĪ

0.16

udas

0.16

uda

0.15

.PO

0.15

inki

0.14

andi

0.14

vester

0.14

plr

0.14

 @{↵

0.14

ä¸Ģäºº

0.14

Activations Density 0.075%

words and phrases associated with violence and physical harm

No Comments

No Known Activations