INDEX

Explanations

terms associated with code modification and manipulation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

ibir

-0.15

Ð½ÐµÐ½

-0.15

pob

-0.15

flo

-0.14

oriously

-0.14

ÑĢÐ°Ð´Ð¸

-0.14

edii

-0.14

kenin

-0.14

ÎĿ

-0.14

(çģ«

-0.14

POSITIVE LOGITS

 tens

0.16

à¸Ļà¸²à¸¡

0.15

effective

0.15

-effect

0.14

olem

0.14

éģ¸æīĭ

0.14

 alter

0.14

ffect

0.14

 directly

0.14

Activations Density 0.198%

terms associated with code modification and manipulation

No Comments

No Known Activations