INDEX

Explanations

terms related to errors, flaws, and failures in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ilon

-0.15

vanished

-0.15

quirer

-0.15

 Ð½Ð°Ð´Ð°

-0.14

ismatch

-0.14

ErrorMsg

-0.14

ÅĤu

-0.13

held

-0.13

ILON

-0.13

acket

-0.13

POSITIVE LOGITS

/error

0.27

/errors

0.26

 committed

0.23

iveness

0.21

/problem

0.20

cies

0.19

Occurred

0.19

/Error

0.19

/loose

0.18

 tolerance

0.18

Activations Density 0.130%

terms related to errors, flaws, and failures in various contexts

No Comments

No Known Activations