INDEX

Explanations

numerical references, such as measurements, quantities, and identifiers, within a document

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

WithValue

-0.15

RO

-0.15

atable

-0.14

_RB

-0.14

-R

-0.14

MM

-0.14

ROL

-0.14

gere

-0.14

POSITIVE LOGITS

eus

0.21

Ð°Ð½Ðº

0.17

ãĥ¼ãĥĩ

0.16

DEX

0.15

 Lager

0.15

¤í

0.15

ÐµÐ´Ð¸

0.15

IFF

0.15

lex

0.15

anki

0.14

Activations Density 0.049%

numerical references, such as measurements, quantities, and identifiers, within a document

No Comments

No Known Activations