INDEX

Explanations

numerical references or metrics related to specific topics or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 affected

-0.17

Ī

-0.17

iken

-0.16

affected

-0.15

uvw

-0.14

 pulses

-0.14

ccione

-0.14

zia

-0.14

Hud

-0.14

POSITIVE LOGITS

orners

0.16

 circum

0.15

çī

0.14

uji

0.14

okens

0.14

 perimeter

0.14

 Circuit

0.14

 circuit

0.14

ungan

0.14

 Seal

0.13

Activations Density 0.025%

numerical references or metrics related to specific topics or entities

No Comments

No Known Activations