INDEX

Explanations

sections or references to appendices, notes, or listings in documents

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

UMB

-0.16

okens

-0.15

 char

-0.15

lak

-0.15

umb

-0.14

emu

-0.14

.repeat

-0.14

opal

-0.14

anced

-0.14

Raw

-0.14

POSITIVE LOGITS

fft

0.17

zyst

0.16

ÐºÐ¾Ð·

0.16

IColor

0.15

ãĤ¤ãĤ¹

0.15

Duy

0.14

plx

0.14

à¥Ĥà¤¬

0.14

dzi

0.14

vim

0.14

Activations Density 0.046%

sections or references to appendices, notes, or listings in documents

No Comments

No Known Activations