INDEX

Explanations

capitalized names and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

lemetry

-0.18

ÙĬØªÙĬ

-0.16

 YÃ¶n

-0.16

loub

-0.15

ptal

-0.15

ovna

-0.15

eniable

-0.15

enheim

-0.15

ctrine

-0.15

apan

-0.14

POSITIVE LOGITS

ondon

0.16

à¤Łà¤¨

0.14

reat

0.14

 Stevens

0.14

-m

0.14

jig

0.13

ear

0.13

 Milano

0.13

atab

0.13

eh

0.13

Activations Density 0.111%

capitalized names and titles

No Comments

No Known Activations