INDEX

Explanations

references to legal terminology and judicial context

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

ãĥ©ãĤ¤ãĥĪ

-0.18

ë¥

-0.16

à¤Ĭ

-0.16

apur

-0.16

ino

-0.15

à¹ĥà¸Ī

-0.15

ìĿ¸íĬ¸

-0.15

rai

-0.14

ipo

-0.14

olla

-0.14

POSITIVE LOGITS

 pack

0.17

pack

0.17

ais

0.16

orz

0.15

usz

0.15

 Pack

0.15

pak

0.15

Fet

0.14

hus

0.14

avers

0.14

Activations Density 0.029%

references to legal terminology and judicial context

No Comments

No Known Activations