INDEX

Explanations

recurrent phrases indicating certainty or possibility in statements

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

kaar

-0.17

 jadx

-0.15

mitt

-0.15

à¸´à¸į

-0.15

ãİ¡

-0.14

vailability

-0.14

KER

-0.14

ÑĮÐ²

-0.13

ãģ£ãģ¨

-0.13

WARDED

-0.13

POSITIVE LOGITS

 follows

0.24

 follow

0.23

 suff

0.20

 trans

0.20

 should

0.19

 must

0.19

 worth

0.18

follow

0.18

 Follow

0.18

 Suff

0.17

Activations Density 0.110%

recurrent phrases indicating certainty or possibility in statements

No Comments

No Known Activations