INDEX

Explanations

references to geographic locations and landmarks

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

.obtain

-0.17

antee

-0.15

ãĥ³ãĥĨãĤ£

-0.15

Ð¾Ð¶

-0.14

UNICATION

-0.14

infeld

-0.14

chy

-0.14

wang

-0.14

 leve

-0.14

Î±Î½

-0.14

POSITIVE LOGITS

 Bermuda

0.37

 Berm

0.34

 Hamilton

0.22

erm

0.19

 Throne

0.18

Hamilton

0.18

 shorts

0.18

BER

0.17

 Pemb

0.17

ado

0.17

Activations Density 0.016%

references to geographic locations and landmarks

No Comments

No Known Activations