INDEX

Explanations

references to religious text and concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

abad

-0.15

 Hussein

-0.15

ella

-0.15

-0.14

alg

-0.14

 Islam

-0.14

ammers

-0.14

Î¼ÏĮ

-0.14

avel

-0.14

 trif

-0.14

POSITIVE LOGITS

Sur

0.22

-sur

0.21

sur

0.21

Sur

0.19

 Ø§ÙĦØ¨ÙĦØ¯

0.18

SUR

0.17

SUR

0.17

sur

0.16

'gc

0.16

ÙģØµÙĦ

0.16

Activations Density 0.015%

references to religious text and concepts

No Comments

No Known Activations

references to religious text and concepts

No Comments

No Known Activations