INDEX

Explanations

expressions of potential, capability, or state of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

èĥ½å¤Ł

-0.18

being

-0.17

 being

-0.17

ABLE

-0.17

èĥ½

-0.16

 ability

-0.16

-being

-0.16

 able

-0.15

 Ability

-0.14

POSITIVE LOGITS

 easily

0.30

 traced

0.27

 Easily

0.22

 anything

0.22

 liken

0.21

anything

0.21

 either

0.20

 anywhere

0.20

found

0.19

 safely

0.19

Activations Density 0.160%

expressions of potential, capability, or state of being

No Comments

No Known Activations