INDEX

Explanations

instances of specific symbols or characters, indicating a focus on non-verbal content or formatting elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

vd

-0.17

 cooperating

-0.15

 Partner

-0.14

yntax

-0.14

Partner

-0.14

 Cooperation

-0.14

 registrazione

-0.14

åĲĪä½ľ

-0.14

 eskort

-0.14

 consort

-0.14

POSITIVE LOGITS

 responsibility

0.21

 oversee

0.21

 responsibilities

0.20

 overseeing

0.20

 oversees

0.19

 duties

0.18

 oversight

0.18

 assignment

0.18

 charge

0.18

 overse

0.17

Activations Density 0.010%

instances of specific symbols or characters, indicating a focus on non-verbal content or formatting elements

No Comments

No Known Activations