INDEX

Explanations

references to prior information or statements made in the text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ÑĸÐ½Ð°

-0.17

eree

-0.16

Ð¸Ð½Ð°

-0.15

ina

-0.14

 Ø®Ø§Ø±Ø¬ÙĬØ©

-0.14

il

-0.14

 siguientes

-0.14

arth

-0.13

uner

-0.13

zac

-0.13

POSITIVE LOGITS

 above

0.68

above

0.59

 Above

0.55

Above

0.54

 ABOVE

0.51

_above

0.46

 Ð²ÑĭÑĪÐµ

0.46

ä¸Ĭ

0.39

 vÃ½Å¡e

0.38

 earlier

0.38

Activations Density 0.125%

references to prior information or statements made in the text

No Comments

No Known Activations