INDEX

Explanations

occurrences of specific characters or formatting commonly associated with section headers or titles in text

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

pie

-0.16

ASCII

-0.16

ICLE

-0.16

Fcn

-0.15

.tie

-0.15

achts

-0.15

 Pier

-0.15

-bot

-0.15

tie

-0.15

Tony

-0.15

POSITIVE LOGITS

Rev

0.22

rev

0.20

Rev

0.19

 Reverse

0.19

 revolution

0.18

 revolving

0.18

 reverse

0.17

 Revolutionary

0.17

reverse

0.17

rev

0.17

Activations Density 0.006%

occurrences of specific characters or formatting commonly associated with section headers or titles in text

No Comments

No Known Activations