INDEX

Explanations

phrases indicating possible problems or issues with proposed solutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

ÎºÎ¹

-0.18

 Ã§Ã¼nkÃ¼

-0.17

 Ø²ÛĮØ±Ø§

-0.16

ONEY

-0.16

but

-0.14

 thereby

-0.14

 takÅ¾e

-0.14

 Erotik

-0.14

 hete

-0.14

uggy

-0.14

POSITIVE LOGITS

 like

0.19

 unlike

0.18

 along

0.17

 once

0.17

along

0.16

 though

0.16

 meanwhile

0.16

ï¼īãģ¯

0.15

elt

0.15

gether

0.14

Activations Density 0.142%

phrases indicating possible problems or issues with proposed solutions

No Comments

No Known Activations