INDEX

Explanations

references to government spending, accountability, and the effectiveness of funding in education and infrastructure

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 reimburse

-0.16

 reimb

-0.15

 indemn

-0.15

dag

-0.14

ambre

-0.14

uhe

-0.14

Ð¸ÑĤÐ¾ÑĢ

-0.14

.sessions

-0.13

Ð¼Ð¾ÑĢ

-0.13

POSITIVE LOGITS

use

0.36

 uses

0.36

 Uses

0.30

Use

0.30

 spending

0.29

 spends

0.28

 spend

0.28

Uses

0.27

Use

0.27

use

0.26

Activations Density 0.178%

references to government spending, accountability, and the effectiveness of funding in education and infrastructure

No Comments

No Known Activations