INDEX

Explanations

programming syntax, particularly in relation to regular expressions and identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

odom

-0.18

zym

-0.15

swer

-0.15

extra

-0.15

zim

-0.14

odynam

-0.14

bay

-0.14

bish

-0.14

rix

-0.14

phyl

-0.14

POSITIVE LOGITS

ws

0.17

 bang

0.16

ikel

0.16

 optional

0.16

Kle

0.16

terminated

0.15

 identification

0.15

 Seymour

0.15

 Binder

0.15

WS

0.15

Activations Density 0.022%

programming syntax, particularly in relation to regular expressions and identifiers

No Comments

No Known Activations