INDEX

Explanations

tokens or characters with high frequency or specific formatting, potentially indicative of programming or code elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

.stamp

-0.17

errat

-0.17

vetica

-0.16

olini

-0.16

ivent

-0.14

sab

-0.14

ream

-0.13

Sab

-0.13

steder

-0.13

POSITIVE LOGITS

 point

0.25

 Point

0.25

che

0.22

point

0.22

-point

0.21

Point

0.21

 position

0.20

pop

0.20

é»ŀ

0.19

che

0.19

Activations Density 0.007%

tokens or characters with high frequency or specific formatting, potentially indicative of programming or code elements

No Comments

No Known Activations