INDEX

Explanations

negative or non-positive terms related to mathematical or physical concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

reff

-0.18

 Hlav

-0.15

vÄĽÅĻ

-0.14

tmpl

-0.14

alist

-0.14

*)_

-0.14

 Moran

-0.14

onis

-0.14

 Electro

-0.13

POSITIVE LOGITS

pNet

0.19

ìĺ

0.16

oriously

0.15

ancode

0.15

olk

0.15

vais

0.14

¶Į

0.14

ÐµÑĤÐµÐ»ÑĮ

0.14

rej

0.14

eme

0.14

Activations Density 0.021%

negative or non-positive terms related to mathematical or physical concepts

No Comments

No Known Activations