INDEX

Explanations

phrases that start with "For" indicating examples or reasons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ãĥĭãĥĥãĤ¯

-0.16

owie

-0.14

ctype

-0.14

parm

-0.14

imals

-0.14

 Ø³Ø®

-0.13

uclear

-0.13

inding

-0.13

vang

-0.13

jection

-0.13

POSITIVE LOGITS

 ná»¯a

0.16

ster

0.14

Ø³ØªØ±

0.14

 intents

0.14

ess

0.14

adam

0.14

-UA

0.14

æĿ¥è¯´

0.14

 instance

0.14

Argb

0.13

Activations Density 0.080%

phrases that start with "For" indicating examples or reasons

No Comments

No Known Activations