INDEX

Explanations

references to locations or geographical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.19.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

rint

-0.16

izza

-0.15

ills

-0.14

hou

-0.14

Ø±ÙĬÙģ

-0.14

conda

-0.13

ulton

-0.13

zier

-0.13

upa

-0.13

lij

-0.13

POSITIVE LOGITS

 which

0.43

 whose

0.37

which

0.34

 ÙĪØ§ÙĦØªÙĬ

0.31

whose

0.30

 WHICH

0.27

 ÐºÐ¾ÑĤÐ¾ÑĢÐ°Ñı

0.27

 whom

0.26

 ÐºÐ¾ÑĤÐ¾ÑĢÑĭÐ¹

0.25

 wich

0.25

Activations Density 0.262%

references to locations or geographical features

No Comments

No Known Activations