INDEX

Explanations

references to television shows and their ratings

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

ypass

-0.15

xy

-0.15

YTE

-0.15

å¥ĳ

-0.14

idon

-0.14

isson

-0.14

 succ

-0.14

 wise

-0.13

 Succ

-0.13

phy

-0.13

POSITIVE LOGITS

ãĥ³ãĥķ

0.18

cesso

0.14

otime

0.14

ksam

0.14

addy

0.14

à¹ģà¸«à¸¥

0.14

ackbar

0.14

 Flying

0.13

specified

0.13

 Hank

0.13

Activations Density 0.100%

references to television shows and their ratings

No Comments

No Known Activations

references to television shows and their ratings

No Comments

No Known Activations