INDEX

Explanations

references to regulatory or disciplinary processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

clarations

-0.16

 plaintiffs

-0.15

 criminals

-0.14

ekl

-0.14

ekk

-0.14

assis

-0.14

 plur

-0.14

iap

-0.14

OUNT

-0.14

 plaintiff

-0.14

POSITIVE LOGITS

 hearing

0.44

 Hearing

0.41

 Hear

0.34

 hearings

0.32

 hear

0.28

 heard

0.27

 administrative

0.26

 hears

0.25

hear

0.24

heard

0.23

Activations Density 0.122%

references to regulatory or disciplinary processes

No Comments

No Known Activations