INDEX

Explanations

phrases indicating personal identification and societal labels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

Introduced

-0.16

.requests

-0.14

AXB

-0.14

veal

-0.14

áº½

-0.13

undai

-0.13

 ComVisible

-0.13

Ã¡la

-0.13

steller

-0.13

roduced

-0.13

POSITIVE LOGITS

 thought

0.78

 believed

0.67

 said

0.62

thought

0.62

 Thought

0.59

Thought

0.50

said

0.49

 alleged

0.45

 reported

0.43

bel

0.40

Activations Density 0.254%

phrases indicating personal identification and societal labels

No Comments

No Known Activations