INDEX

Explanations

references to halting processes, suspensions, or interruptions in progress

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ationally

-0.15

-0.14

ints

-0.14

 removeFromSuperview

-0.14

disposed

-0.13

ign

-0.13

bere

-0.13

æ¡

-0.13

 Podesta

-0.13

ÅĻet

-0.13

POSITIVE LOGITS

 suspend

0.40

 halt

0.39

 susp

0.38

 suspended

0.38

 halted

0.37

 suspension

0.37

 stop

0.37

 stopped

0.36

 paused

0.36

 pause

0.35

Activations Density 0.188%

references to halting processes, suspensions, or interruptions in progress

No Comments

No Known Activations