INDEX

Explanations

references to short durations or timeframes

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

ustum

-0.16

lical

-0.15

aten

-0.14

hra

-0.14

 Schiff

-0.14

ICI

-0.14

authenticated

-0.13

ulumi

-0.13

ic

-0.13

(strtolower

-0.13

POSITIVE LOGITS

-term

0.24

ening

0.21

-lived

0.21

listed

0.21

ened

0.20

(er

0.18

/tiny

0.18

term

0.18

comings

0.17

-short

0.17

Activations Density 0.027%

references to short durations or timeframes

No Comments

No Known Activations