INDEX

Explanations

references to walls and related surfaces or coverings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

åłĤ

-0.17

eka

-0.17

ìľ¨

-0.16

OrCreate

-0.16

elli

-0.15

serter

-0.15

_MR

-0.15

ement

-0.14

udes

-0.14

Î»Î¹ÎºÎ¬

-0.14

POSITIVE LOGITS

abies

0.27

-mounted

0.26

aby

0.26

å£ģ

0.23

ace

0.22

-wall

0.20

 wall

0.20

 Wall

0.20

papers

0.20

/window

0.20

Activations Density 0.030%

references to walls and related surfaces or coverings

No Comments

No Known Activations