INDEX

Explanations

negative assertions or claims about existence and value

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

æ¢

-0.16

quip

-0.15

bes

-0.14

ugu

-0.14

fell

-0.14

lw

-0.14

 besides

-0.13

pers

-0.13

PHA

-0.13

olet

-0.13

POSITIVE LOGITS

ehr

0.17

gesch

0.16

nor

0.15

 cins

0.15

anh

0.15

bout

0.14

holding

0.14

ãĤ¹ãĥŀ

0.14

ebra

0.14

suma

0.14

Activations Density 0.406%

negative assertions or claims about existence and value

No Comments

No Known Activations