INDEX

Explanations

phrases indicating actions or commands related to accessing information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

tam

-0.16

ennen

-0.15

Ì§

-0.15

INVAL

-0.14

uos

-0.14

à¥ĭà¤§

-0.14

adro

-0.13

cÃ¼

-0.13

ÑĩÐ¸

-0.13

Hierarchy

-0.13

POSITIVE LOGITS

her

0.45

 hear

0.44

-h

0.42

hear

0.31

HER

0.31

 hete

0.30

Âłh

0.29

_h

0.27

Aqu

0.26

her

0.26

Activations Density 0.123%

phrases indicating actions or commands related to accessing information

No Comments

No Known Activations