INDEX

Explanations

instances of product or object identification

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

Traversal

-0.18

egin

-0.18

ulse

-0.16

estival

-0.15

pekt

-0.14

 Bennett

-0.14

uet

-0.14

pecial

-0.14

ÑĢÐ¸Ð·

-0.14

squ

-0.13

POSITIVE LOGITS

oppable

0.16

emale

0.15

è·

0.14

ë¶

0.14

 åľ

0.14

 alcoholic

0.14

olem

0.14

anki

0.14

ERRU

0.14

lion

0.14

Activations Density 0.005%

instances of product or object identification

No Comments

No Known Activations