INDEX

Explanations

references to passwords and related security terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

pping

-0.16

oleans

-0.15

etro

-0.15

 Presidency

-0.14

OfYear

-0.14

ë§ī

-0.14

 presidency

-0.14

ucher

-0.14

rowser

-0.14

infeld

-0.14

POSITIVE LOGITS

Strength

0.24

 strength

0.23

-reset

0.22

_strength

0.22

 Strength

0.21

/pass

0.20

strength

0.20

 reset

0.20

 Reset

0.19

_reset

0.18

Activations Density 0.009%

references to passwords and related security terms

No Comments

No Known Activations

references to passwords and related security terms

No Comments

No Known Activations