INDEX

Explanations

occurrences of "news" and related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

_news

-0.17

iggs

-0.17

neau

-0.17

ëī´ìĬ¤

-0.17

toolbox

-0.16

asco

-0.16

ept

-0.15

ews

-0.15

ting

-0.15

ping

-0.15

POSITIVE LOGITS

room

0.37

letters

0.31

flash

0.30

stand

0.29

reader

0.28

stands

0.28

rooms

0.27

feed

0.25

agent

0.24

ROOM

0.24

Activations Density 0.043%

occurrences of "news" and related concepts

No Comments

No Known Activations