INDEX

Explanations

terms related to features or characteristics of products or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

-0.23

sWith

-0.21

sav

-0.18

sit

-0.18

feas

-0.17

ners

-0.16

sian

-0.15

_feature

-0.15

å¾Ĵ

-0.15

gow

-0.15

POSITIVE LOGITS

tte

0.39

 prominently

0.30

less

0.20

-length

0.19

-rich

0.19

ettes

0.18

etro

0.18

et

0.17

-packed

0.17

etto

0.17

Activations Density 0.036%

terms related to features or characteristics of products or subjects

No Comments

No Known Activations