INDEX

Explanations

terms related to fixing and repair processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

hill

-0.16

ilet

-0.15

lig

-0.15

alu

-0.15

ERSHEY

-0.15

loor

-0.15

cipher

-0.14

andin

-0.14

actor

-0.14

veau

-0.14

POSITIVE LOGITS

tures

0.24

TURE

0.21

(es

0.20

gerald

0.18

ADDE

0.18

.fix

0.17

-fixed

0.17

Fix

0.17

emean

0.16

xed

0.16

Activations Density 0.021%

terms related to fixing and repair processes

No Comments

No Known Activations