INDEX

Explanations

names associated with individuals in news articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

 Gregg

-0.17

olla

-0.16

amac

-0.14

otron

-0.14

zym

-0.14

 till

-0.13

apons

-0.13

ination

-0.13

TB

-0.13

POSITIVE LOGITS

@student

0.16

 Chatt

0.15

CommandEvent

0.15

CJK

0.15

emean

0.14

ãģ®ãĤĤ

0.14

oksen

0.13

ifa

0.13

 Bald

0.13

ÑĢÐ°Ð¿

0.13

Activations Density 0.024%

names associated with individuals in news articles

No Comments

No Known Activations