INDEX

Explanations

references to military personnel and symbols related to them

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

RTC

-0.16

 police

-0.15

OPS

-0.15

erton

-0.14

WithValue

-0.14

 Police

-0.14

assis

-0.14

.ud

-0.14

 distraction

-0.13

 cancel

-0.13

POSITIVE LOGITS

POW

0.51

 prisoner

0.49

 captured

0.48

 prisoners

0.46

 captive

0.45

 capture

0.44

 capt

0.44

 Prison

0.43

 captures

0.43

 captivity

0.42

Activations Density 0.079%

references to military personnel and symbols related to them

No Comments

No Known Activations