INDEX

Explanations

variations of the word "dim," particularly in relation to visibility or light

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Ã¶l

-0.17

 Savage

-0.16

Err

-0.16

gar

-0.15

gor

-0.15

ãĤŃãĥ³ãĤ°

-0.15

iversary

-0.14

yny

-0.14

:CGPoint

-0.14

ame

-0.14

POSITIVE LOGITS

ENSION

0.34

inished

0.30

ENSIONS

0.30

ethyl

0.28

ensions

0.26

inish

0.25

ention

0.25

anche

0.24

ension

0.23

orphic

0.23

Activations Density 0.014%

variations of the word "dim," particularly in relation to visibility or light

No Comments

No Known Activations