INDEX

Explanations

specific terms related to chess

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

èĻ«

-0.16

-0.15

ayment

-0.15

 McCart

-0.14

 Nichols

-0.14

ighton

-0.14

theon

-0.13

 mutually

-0.13

auga

-0.13

rates

-0.13

POSITIVE LOGITS

eger

0.16

åħ³

0.15

 duty

0.14

ÑĮÐºÐ¾

0.14

 Duty

0.14

éĸ¢

0.14

ioc

0.14

auled

0.13

afa

0.13

ges

0.13

Activations Density 0.000%

specific terms related to chess

No Comments

No Known Activations