INDEX

Explanations

references to structural features and connections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

Gee

-0.15

inness

-0.15

eland

-0.15

 truncate

-0.14

ãĥĥãĥĦ

-0.14

utar

-0.14

 Papers

-0.14

reek

-0.14

-0.13

ubic

-0.13

POSITIVE LOGITS

é¬¼

0.18

Ð»ÑİÑĩÐ°

0.15

chandle

0.15

ÏħÎ³

0.15

apture

0.14

fra

0.14

arket

0.14

ØµØ¯

0.14

lag

0.14

Activations Density 0.039%

references to structural features and connections

No Comments

No Known Activations