INDEX

Explanations

geographical locations and related place names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Minnesota

-0.16

orra

-0.15

Enc

-0.15

 Manitoba

-0.15

witch

-0.14

MN

-0.14

Enc

-0.14

 root

-0.14

ziej

-0.14

Minnesota

-0.14

POSITIVE LOGITS

 Charleston

0.41

 Clemson

0.35

SC

0.35

 Greenville

0.32

 Carolina

0.32

 South

0.31

 Spartan

0.30

 SOUTH

0.29

(SC

0.28

South

0.28

Activations Density 0.083%

geographical locations and related place names

No Comments

No Known Activations