INDEX

Explanations

references to governmental and legislative actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

entions

-0.16

 span

-0.15

ÑĥÑĩÐ°Ñģ

-0.15

ÎºÎŃ

-0.14

 Village

-0.14

span

-0.14

 District

-0.14

 village

-0.14

rink

-0.13

mk

-0.13

POSITIVE LOGITS

asury

0.18

 Admir

0.18

Office

0.18

 Treasury

0.17

 Office

0.17

 Inner

0.17

 mint

0.17

-office

0.17

 office

0.17

 kancel

0.16

Activations Density 0.080%

references to governmental and legislative actions

No Comments

No Known Activations