INDEX

Explanations

actions and verbs indicating initiative or response

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

rej

-0.17

ulle

-0.15

undle

-0.15

apon

-0.15

ÐµÑĢÐ¶

-0.15

rej

-0.14

viÄį

-0.14

emouth

-0.14

APON

-0.14

.Accessible

-0.14

POSITIVE LOGITS

 action

0.35

 Action

0.28

action

0.26

 steps

0.26

.action

0.26

 actions

0.26

Action

0.25

_action

0.25

 ACTION

0.23

-action

0.23

Activations Density 0.044%

actions and verbs indicating initiative or response

No Comments

No Known Activations