INDEX

Explanations

links to external resources 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

 counterpart

-0.15

 correspondent

-0.14

oose

-0.14

 correspond

-0.14

.fromCharCode

-0.13

 Fancy

-0.13

POSITIVE LOGITS

uali

0.16

erez

0.15

Executor

0.15

ancellable

0.15

amps

0.14

ÄŁer

0.14

acos

0.14

 Andersen

0.14

variant

0.13

'gc

0.13

Activations Density 1.046%

links to external resources and articles

No Comments

No Known Activations