INDEX

Explanations

specific proper nouns, including names and titles

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

ov

-0.17

val

-0.17

åħĥ

-0.16

aval

-0.16

eline

-0.15

 Walton

-0.15

rees

-0.14

vala

-0.14

zeug

-0.14

Luc

-0.14

POSITIVE LOGITS

redo

0.15

atorial

0.15

anken

0.15

pll

0.14

arin

0.14

ãĤ¹ãĥ¬

0.14

anity

0.14

ytt

0.14

 Arts

0.14

gif

0.14

Activations Density 0.054%

specific proper nouns, including names and titles

No Comments

No Known Activations