INDEX

Explanations

high-frequency, non-verbal symbols or characters often associated with code or technical specifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

atore

-0.17

eres

-0.17

Unused

-0.15

èĮĥ

-0.15

Ã©rc

-0.15

anky

-0.15

yne

-0.14

Ãºp

-0.14

æĳĺ

-0.14

otu

-0.14

POSITIVE LOGITS

asca

0.22

auty

0.19

burg

0.15

Ð¾ÑĢÐ¾

0.15

meli

0.15

igg

0.14

bih

0.14

à¤Ĥà¤Ń

0.14

iard

0.14

 Seks

0.14

Activations Density 0.086%

high-frequency, non-verbal symbols or characters often associated with code or technical specifications

No Comments

No Known Activations