INDEX

Explanations

references to specific individuals or characters identified by the letter "P."

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

ure

-0.19

las

-0.17

assel

-0.17

ì§ĵ

-0.17

EMENT

-0.17

ague

-0.15

aste

-0.15

(.)

-0.15

ëł¬

-0.15

à¤¾à¤¡

-0.15

POSITIVE LOGITS

 Duty

0.17

oy

0.16

 sign

0.15

ledon

0.15

dem

0.15

 duty

0.15

um

0.14

onn

0.14

cola

0.14

Activations Density 0.048%

references to specific individuals or characters identified by the letter "P."

No Comments

No Known Activations