INDEX

Explanations

references to suspects and arrests in criminal cases

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

-valu

-0.15

 DISPATCH

-0.15

orden

-0.14

ioc

-0.14

à¸«à¸¡

-0.14

 Victims

-0.14

ffect

-0.14

 Assass

-0.14

icens

-0.14

anus

-0.13

POSITIVE LOGITS

 wanted

0.30

 suspects

0.28

 charged

0.26

 suspect

0.26

 suspected

0.25

 Wanted

0.25

sus

0.24

wanted

0.24

 appreh

0.24

 accused

0.23

Activations Density 0.121%

references to suspects and arrests in criminal cases

No Comments

No Known Activations

references to suspects and arrests in criminal cases

No Comments

No Known Activations