INDEX

Explanations

references to legal principles or terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

rlen

-0.20

icolon

-0.16

achsen

-0.15

_python

-0.15

 Journal

-0.14

mic

-0.14

 Tribe

-0.14

ouri

-0.14

dump

-0.14

 journal

-0.14

POSITIVE LOGITS

Aqu

0.27

 Scot

0.26

 Dominican

0.25

 gloss

0.24

Aqu

0.23

sch

0.21

 Dante

0.21

Fri

0.21

 Gros

0.21

 medieval

0.20

Activations Density 0.039%

references to legal principles or terminology

No Comments

No Known Activations