INDEX

Explanations

phrases indicating conflict or contradiction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

iffer

-0.17

 Äĳá»Ŀi

-0.15

urum

-0.15

mez

-0.14

dney

-0.14

ifer

-0.14

 æĢ§

-0.13

unch

-0.13

è¥

-0.13

emale

-0.13

POSITIVE LOGITS

EVT

0.16

_mk

0.15

ê¸°ê´Ģ

0.14

jvu

0.14

 others

0.14

iske

0.14

bjerg

0.14

Others

0.13

 Others

0.13

Activations Density 0.215%

phrases indicating conflict or contradiction

No Comments

No Known Activations