INDEX

Explanations

phrases that emphasize maximizing or optimizing outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

 Scatter

-0.15

 scatter

-0.15

ograd

-0.15

éĻ£

-0.14

abay

-0.14

InThe

-0.14

agu

-0.14

_UT

-0.14

Sampler

-0.13

aul

-0.13

POSITIVE LOGITS

out

0.36

 extraction

0.33

 extract

0.32

 Extract

0.31

 Extraction

0.30

 extracted

0.30

 extracting

0.28

Extract

0.27

 extracts

0.27

 squeezing

0.27

Activations Density 0.089%

phrases that emphasize maximizing or optimizing outcomes

No Comments

No Known Activations