INDEX

Explanations

programming structures and control flow statements, particularly loops

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ffa

-0.16

Ã´i

-0.15

Ð°ÑĢÐ°Ð¼

-0.15

esen

-0.15

aton

-0.15

enstein

-0.14

ilent

-0.14

tile

-0.14

ÑĭÑģ

-0.14

decorate

-0.14

POSITIVE LOGITS

int

0.20

æ£

0.17

 purposes

0.17

ero

0.16

cken

0.16

_,

0.16

let

0.16

 sake

0.16

zÄĻ

0.16

iero

0.15

Activations Density 0.036%

programming structures and control flow statements, particularly loops

No Comments

No Known Activations