INDEX

Explanations

elements related to bibliographic information and citations in academic texts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Simpson

-0.16

aring

-0.15

Ð°ÑĤÐ¾Ðº

-0.14

rescia

-0.14

 sake

-0.14

ForResource

-0.14

entai

-0.14

arem

-0.14

rex

-0.14

resenter

-0.14

POSITIVE LOGITS

idos

0.17

coin

0.15

ãĥ©ãĤ¯

0.14

 upstream

0.14

idis

0.14

eyJ

0.14

 Heller

0.14

marshall

0.14

ido

0.14

ointment

0.14

Activations Density 0.002%

elements related to bibliographic information and citations in academic texts

No Comments

No Known Activations