INDEX

Explanations

connections or associations indicated by the conjunction "and"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 3-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L3R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.3.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

igu

-0.15

ophilia

-0.15

-0.14

agh

-0.14

æ£Ĵ

-0.14

ÑĢÑĥÐ½

-0.14

wn

-0.13

chemist

-0.13

Whe

-0.13

POSITIVE LOGITS

kus

0.16

tabpanel

0.15

alink

0.15

oyer

0.14

wie

0.14

allen

0.14

deck

0.14

idak

0.14

agne

0.13

Activations Density 0.115%

connections or associations indicated by the conjunction "and"

No Comments

No Known Activations