INDEX

Explanations

conjunctions and articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

emat

-0.16

gere

-0.15

eco

-0.14

tera

-0.14

emma

-0.14

ead

-0.14

Ð»Ð°Ð±

-0.14

.sdk

-0.13

ibir

-0.13

olest

-0.13

POSITIVE LOGITS

/or

0.19

ies

0.16

/of

0.15

onso

0.15

 Bols

0.14

Ø§Ø¨Ø±

0.14

lt

0.14

å¼ı

0.14

 çł

0.13

rss

0.13

Activations Density 0.288%

conjunctions and articles

No Comments

No Known Activations