INDEX

Explanations

references to crime-related media and authors

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

ãĥ³ãĤ¬

-0.17

.scalablytyped

-0.16

 stripslashes

-0.15

Ð°ÑĢÑĩ

-0.15

HX

-0.15

geist

-0.14

ç¦

-0.14

bler

-0.14

lÃ©m

-0.14

imizer

-0.14

POSITIVE LOGITS

DS

0.17

 recip

0.16

pei

0.15

rott

0.15

DI

0.15

Nga

0.14

ÏĦÎ¿Îº

0.14

 suspense

0.14

è¿ª

0.14

 Virgin

0.14

Activations Density 0.025%

references to crime-related media and authors

No Comments

No Known Activations