INDEX

Explanations

references to time

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

aby

-0.18

ostat

-0.18

stadt

-0.17

 Platt

-0.16

terra

-0.16

åŀ

-0.15

iker

-0.15

upp

-0.15

Ã´le

-0.15

alla

-0.14

POSITIVE LOGITS

bomb

0.22

 Warner

0.22

 honored

0.22

 capsule

0.21

 honoured

0.21

 dilation

0.20

hon

0.20

 warp

0.20

Machine

0.20

_machine

0.19

Activations Density 0.045%

references to time

No Comments

No Known Activations