INDEX

Explanations

references to state and government entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 states

-0.20

_states

-0.18

 States

-0.17

States

-0.17

asser

-0.17

çĬ¶æĢģ

-0.17

 ÑģÐ¾ÑģÑĤÐ¾Ñı

-0.16

InstanceState

-0.15

ãĤ£

-0.15

 estados

-0.15

POSITIVE LOGITS

wide

0.39

-wide

0.30

/local

0.29

Wide

0.28

 wide

0.28

hood

0.27

house

0.26

 Wide

0.24

tro

0.23

-level

0.21

Activations Density 0.039%

references to state and government entities

No Comments

No Known Activations