INDEX

Explanations

terms related to deception and manipulation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

lient

-0.16

rott

-0.15

rophe

-0.15

å¹¹

-0.15

eming

-0.14

æ¥

-0.14

fsp

-0.14

ikip

-0.14

ackers

-0.14

umi

-0.14

POSITIVE LOGITS

ero

0.17

sters

0.17

 tricks

0.16

eries

0.16

ery

0.16

ÑĥÐ»Ð¸

0.15

&T

0.15

lez

0.15

ERY

0.14

 Grat

0.14

Activations Density 0.167%

terms related to deception and manipulation

No Comments

No Known Activations