INDEX

Explanations

references to youthful mistakes and themes of accountability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 perhaps

-0.19

 indeed

-0.19

 folks

-0.18

 Perhaps

-0.17

perhaps

-0.17

 Indeed

-0.16

Perhaps

-0.16

 simply

-0.16

 herein

-0.15

 colleague

-0.15

POSITIVE LOGITS

 stuff

0.17

 kind

0.15

 haha

0.15

 weird

0.14

éľ

0.14

ITERAL

0.14

.animations

0.14

 weir

0.14

gross

0.14

icer

0.14

Activations Density 2.212%

references to youthful mistakes and themes of accountability

No Comments

No Known Activations