INDEX

Explanations

themes related to accountability and consequences for actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

dac

-0.17

 Kendrick

-0.16

è

-0.16

mada

-0.15

gii

-0.15

alarda

-0.15

uxt

-0.15

pire

-0.15

erview

-0.14

unicorn

-0.14

POSITIVE LOGITS

 consequences

0.58

 consequence

0.52

 repercussions

0.36

 accountability

0.29

 reperc

0.29

 punishment

0.29

 penalty

0.28

 Accountability

0.25

 ramifications

0.25

 consequential

0.24

Activations Density 0.219%

themes related to accountability and consequences for actions

No Comments

No Known Activations