INDEX

Explanations

words and phrases indicating references to previous content or actions

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

blo

-0.17

èķī

-0.16

 ë¶Ħ

-0.15

ä¿Ŀ

-0.15

anchor

-0.14

 Gund

-0.14

@$_

-0.14

 è±

-0.14

orthand

-0.14

ombine

-0.14

POSITIVE LOGITS

",__

0.16

izar

0.15

ENA

0.15

eti

0.14

EDA

0.14

 Learned

0.14

Esp

0.14

git

0.14

mat

0.14

ket

0.14

Activations Density 0.001%

words and phrases indicating references to previous content or actions

No Comments

No Known Activations