INDEX

Explanations

references to structure and importance 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 @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 Abdel

-0.15

ENG

-0.14

.geo

-0.14

Fld

-0.14

store

-0.14

AttributeName

-0.14

.getClassName

-0.14

itre

-0.13

bsolute

-0.13

 stub

-0.13

POSITIVE LOGITS

().'/

0.19

nier

0.17

ceu

0.15

">#

0.14

arlo

0.14

 this

0.14

ä½µ

0.14

antz

0.14

 Cant

0.13

 rivers

0.13

Activations Density 0.419%

references to structure and importance in various contexts

No Comments

No Known Activations