INDEX

Explanations

code-related actions, particularly those involving data manipulation and structure management

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

à¸¥à¸°

-0.15

 å®ī

-0.14

CY

-0.14

olf

-0.14

 sobie

-0.14

acock

-0.13

 Gross

-0.13

ãĥ¥ãĥ¼

-0.13

áº¯p

-0.13

ãĥ¼ãĥĹ

-0.13

POSITIVE LOGITS

 Delete

0.19

 Removal

0.18

rm

0.18

 delete

0.17

 removal

0.17

 remove

0.17

 removes

0.17

_DELETE

0.17

 Remove

0.17

(delete

0.16

Activations Density 0.132%

code-related actions, particularly those involving data manipulation and structure management

No Comments

No Known Activations