INDEX

Explanations

the word "of"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

oir

-0.16

GLE

-0.16

assin

-0.15

\grid

-0.15

 NUIT

-0.15

>window

-0.15

enticated

-0.15

assis

-0.15

jie

-0.14

á»Ļp

-0.14

POSITIVE LOGITS

 Hipp

0.15

engo

0.14

ucas

0.14

 Achilles

0.14

mente

0.14

 dialect

0.14

rosse

0.14

aley

0.14

uer

0.13

ath

0.13

Activations Density 0.006%

the word "of"

No Comments

No Known Activations