INDEX

Explanations

words related to breaks or pauses in activities

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

abcdef

-0.17

 Mile

-0.15

----</

-0.14

Ú¯Ø§Ø±

-0.13

 Patron

-0.13

 Milk

-0.13

Indexed

-0.13

esson

-0.13

Ð±Ð¸ÑĤ

-0.13

CAF

-0.13

POSITIVE LOGITS

esis

0.18

0.16

ughs

0.15

iÃ¨res

0.15

aal

0.15

okus

0.15

ieg

0.15

Buffers

0.15

å¹³æĪĲ

0.15

Activations Density 0.005%

words related to breaks or pauses in activities

No Comments

No Known Activations