INDEX

Explanations

events and actions that indicate communication or exchanges between characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

duit

-0.15

ï»¿↵↵

-0.14

æĶ¶å½ķ

-0.14

endforeach

-0.13

/jav

-0.13

BCM

-0.13

 ÃľNÄ°

-0.13

 Ð±Ð¾ÐºÑĥ

-0.13

aylor

-0.12

BackPressed

-0.12

POSITIVE LOGITS

.nih

0.14

undle

0.13

aison

0.13

 Farrell

0.13

¢

0.12

oomla

0.12

ulu

0.12

irus

0.12

 Churchill

0.12

Activations Density 0.029%

events and actions that indicate communication or exchanges between characters

No Comments

No Known Activations