INDEX

Explanations

mentions of banks or banking-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

Ã¤d

-0.16

udden

-0.16

adil

-0.15

onest

-0.15

uguay

-0.15

.ul

-0.14

opsis

-0.14

 Boeh

-0.14

comb

-0.14

urnished

-0.14

POSITIVE LOGITS

ruptcy

0.32

rupt

0.25

roll

0.22

rolled

0.21

rolling

0.20

erd

0.20

sters

0.18

er

0.18

ers

0.17

ole

0.17

Activations Density 0.009%

mentions of banks or banking-related terms

No Comments

No Known Activations