INDEX

Explanations

terms indicating news or updates

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

ses

-0.19

attery

-0.18

-ÑĤÐ¾

-0.15

criptions

-0.15

-than

-0.14

Ð¾Ð±ÑĢÐ°Ð·

-0.14

faq

-0.14

rary

-0.13

oga

-0.13

iday

-0.13

POSITIVE LOGITS

WidthSpace

0.18

¤í

0.16

ï¸

0.16

ideographic

0.15

ÅĻÃŃ

0.15

ktop

0.15

íķĺê²Į

0.15

iated

0.14

ØªØ±ÛĮÙĨ

0.14

à¸±à¸ģà¸ª

0.14

Activations Density 0.221%

terms indicating news or updates

No Comments

No Known Activations