INDEX

Explanations

proper nouns, particularly 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 @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

á»ĭ

-0.15

cker

-0.15

ÑĥÐ¶Ðµ

-0.15

wright

-0.14

ÎµÎ¹

-0.14

ubern

-0.14

ampo

-0.14

Ð¾Ð¿ÑĢÐ¾Ñģ

-0.14

ÃŃrk

-0.13

_digest

-0.13

POSITIVE LOGITS

 Rodriguez

0.15

onia

0.14

eria

0.14

san

0.14

_firestore

0.14

 Rodr

0.14

OLA

0.14

uela

0.13

McM

0.13

Activations Density 0.019%

proper nouns, particularly names and titles

No Comments

No Known Activations