INDEX

Explanations

colons used in enumerating categories or tags

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

eland

-0.18

blink

-0.16

 wire

-0.15

wire

-0.15

dorf

-0.15

lant

-0.15

 Guard

-0.14

é£

-0.14

aller

-0.14

 Wire

-0.14

POSITIVE LOGITS

inz

0.15

 Ð¿ÑĢÐ¾Ð¿

0.15

><![

0.15

 DISP

0.15

 DÃ¶n

0.14

çıł

0.14

ç«Ļ

0.14

 Kaynak

0.13

 ÑĢÐ°ÑģÑħÐ¾Ð´

0.13

 Cosby

0.13

Activations Density 0.001%

colons used in enumerating categories or tags

No Comments

No Known Activations