INDEX

Explanations

references to death and mortality

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

icht

-0.16

endoza

-0.16

onders

-0.15

 bloodstream

-0.15

argas

-0.14

Ãµes

-0.14

uler

-0.14

 Dexter

-0.14

osa

-0.14

uest

-0.14

POSITIVE LOGITS

 penalty

0.28

trap

0.23

 Penalty

0.23

ly

0.22

bed

0.21

_penalty

0.21

 sentences

0.20

 sentence

0.20

 thro

0.20

 squads

0.20

Activations Density 0.011%

references to death and mortality

No Comments

No Known Activations