INDEX

Explanations

specific gaming or deck-building references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

Ð¾ÐºÐ¸

-0.15

refund

-0.15

stan

-0.15

 Bend

-0.14

PI

-0.14

jac

-0.14

 Chern

-0.14

 audi

-0.14

Jer

-0.13

ATAB

-0.13

POSITIVE LOGITS

 card

0.33

 cards

0.32

 deck

0.29

deck

0.29

 Cards

0.26

 decks

0.26

 Deck

0.26

Deck

0.25

(Card

0.24

-card

0.24

Activations Density 0.005%

specific gaming or deck-building references

No Comments

No Known Activations