INDEX

Explanations

claims regarding the nature and impacts of objects or concepts, especially those that mislead or present false information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ahl

-0.20

enal

-0.18

lever

-0.16

lÃ¼

-0.15

enthal

-0.15

åĨ

-0.15

-0.14

hea

-0.14

hani

-0.14

eprom

-0.14

POSITIVE LOGITS

aim

0.18

 purpose

0.18

nego

0.17

Aim

0.17

 aims

0.16

aim

0.16

kowski

0.16

aims

0.15

 Morr

0.15

 Purpose

0.15

Activations Density 0.237%

claims regarding the nature and impacts of objects or concepts, especially those that mislead or present false information

No Comments

No Known Activations