INDEX

Explanations

references to patents and associated identification numbers

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

egie

-0.16

iku

-0.16

inch

-0.14

enk

-0.14

kne

-0.14

 getPage

-0.14

Weather

-0.13

Ð½Ð¾ÑĪ

-0.13

 Weather

-0.13

POSITIVE LOGITS

 numberWith

0.19

#:

0.17

.gov

0.16

vur

0.15

cka

0.15

-await

0.15

à¸·à¹ī

0.15

oldur

0.14

 afore

0.14

ucken

0.14

Activations Density 0.066%

references to patents and associated identification numbers

No Comments

No Known Activations