INDEX

Explanations

information related to game mechanics and scoring systems

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

abcdefghijklmnop

-0.17

abcdefghijkl

-0.16

itti

-0.15

oref

-0.15

Fight

-0.14

 Fighting

-0.14

vrier

-0.14

assy

-0.14

 indiv

-0.14

POSITIVE LOGITS

-block

0.17

 timed

0.15

 Conveyor

0.15

 block

0.15

ADO

0.15

 matching

0.15

 Match

0.14

[block

0.14

 conveyor

0.14

 tiles

0.14

Activations Density 0.025%

information related to game mechanics and scoring systems

No Comments

No Known Activations