INDEX

Explanations

references to changes or transformations in various contexts

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

OfType

-0.15

_effects

-0.14

 corrections

-0.14

heten

-0.14

ingroup

-0.13

ture

-0.13

achuset

-0.13

Ð¾Ð´Ð°ÑĢ

-0.13

emodel

-0.13

(Model

-0.13

POSITIVE LOGITS

 gears

0.29

 direction

0.28

 tack

0.28

 focus

0.24

æĸ¹åĲĳ

0.23

 lanes

0.23

 allegiance

0.22

 sides

0.22

 directions

0.22

direction

0.22

Activations Density 0.104%

references to changes or transformations in various contexts

No Comments

No Known Activations