INDEX

Explanations

topics related to crime and violence

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

 Manitoba

-0.17

 Clemson

-0.16

ÙĪÙĬØ³

-0.15

chg

-0.15

 Tulsa

-0.14

izard

-0.14

ync

-0.14

ADE

-0.14

 Manit

-0.14

jiang

-0.14

POSITIVE LOGITS

San

0.70

Bay

0.69

SF

0.65

 Oakland

0.63

SF

0.61

Bay

0.59

San

0.58

bay

0.56

 Berkeley

0.55

bay

0.52

Activations Density 0.258%

topics related to crime and violence

No Comments

No Known Activations