INDEX

Explanations

instances or references to blocks or blocking actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

-0.18

ited

-0.17

ITED

-0.16

ves

-0.16

ikt

-0.15

ooks

-0.15

zion

-0.15

/pm

-0.15

ffects

-0.15

onde

-0.14

POSITIVE LOGITS

buster

0.37

aded

0.35

busters

0.35

ades

0.31

ading

0.29

quote

0.28

chains

0.25

CHAIN

0.23

ade

0.23

upy

0.23

Activations Density 0.023%

instances or references to blocks or blocking actions

No Comments

No Known Activations