INDEX

Explanations

words related to declarations or formal statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

eur

-0.17

 Hurt

-0.16

ÙħØ±

-0.15

oleon

-0.15

 Harm

-0.14

oogle

-0.14

åĭĴ

-0.14

Å¡ti

-0.14

PLUS

-0.14

 Dale

-0.14

POSITIVE LOGITS

utter

0.29

ension

0.26

ination

0.25

ining

0.25

ustering

0.24

ine

0.24

ARATION

0.23

ares

0.23

eration

0.22

ared

0.22

Activations Density 0.005%

words related to declarations or formal statements

No Comments

No Known Activations