INDEX

Explanations

references to mathematical properties, especially those related to integers and sets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ed

-0.16

assador

-0.16

asaki

-0.15

nob

-0.14

ÅĻe

-0.14

áº¯n

-0.14

wood

-0.14

ategorical

-0.14

 Intervention

-0.13

rak

-0.13

POSITIVE LOGITS

swer

0.17

á»§ng

0.16

æ®µ

0.16

ulings

0.16

inati

0.15

ReadStream

0.15

ãģĸ

0.14

zig

0.14

rire

0.14

EXPECT

0.14

Activations Density 0.089%

references to mathematical properties, especially those related to integers and sets

No Comments

No Known Activations