INDEX

Explanations

geographical locations and other place-related references

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

Î±Î½

-0.16

Î»Î»Î¹

-0.15

 Bend

-0.15

.isFile

-0.15

.orm

-0.15

.eulerAngles

-0.15

Ð»Ð¸ÑĤ

-0.15

anje

-0.14

ogs

-0.14

 uten

-0.14

POSITIVE LOGITS

 Oakland

0.37

Oak

0.34

oak

0.27

 Berkeley

0.26

Oak

0.24

Ber

0.22

 Raiders

0.22

0.20

ber

0.20

 Walnut

0.19

Activations Density 0.047%

geographical locations and other place-related references

No Comments

No Known Activations

geographical locations and other place-related references

No Comments

No Known Activations