INDEX

Explanations

references to editing and omitted content in media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ä±klÄ±

-0.16

onom

-0.15

 renewal

-0.14

 improbable

-0.13

ldb

-0.13

 renew

-0.13

Ø¬Ø§Ø±

-0.13

loub

-0.13

Restart

-0.13

APS

-0.13

POSITIVE LOGITS

 cuts

0.28

cut

0.24

cut

0.22

cuts

0.21

 Version

0.21

 version

0.21

Cut

0.21

 alternate

0.19

exc

0.19

 Alternate

0.19

Activations Density 0.084%

references to editing and omitted content in media

No Comments

No Known Activations