INDEX

Explanations

references to studies or citations in research contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

ording

-0.19

caff

-0.16

ooke

-0.15

Ð¾ÑĢÐ´Ð¸Ð½

-0.15

edy

-0.15

/loader

-0.14

bine

-0.14

oten

-0.14

emmel

-0.14

Pont

-0.13

POSITIVE LOGITS

enta

0.17

linky

0.14

orio

0.14

 themselves

0.14

.City

0.13

 tongue

0.13

NullOrEmpty

0.13

abbix

0.12

 reporting

0.12

ACHINE

0.12

Activations Density 0.034%

references to studies or citations in research contexts

No Comments

No Known Activations