INDEX

Explanations

references to people, specifically the name "Smith" and related proper nouns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

iper

-0.15

fen

-0.15

ÑĨÑĸÐ¹

-0.15

 suáº¥t

-0.15

raud

-0.15

rious

-0.14

inqu

-0.14

aways

-0.14

IP

-0.14

acak

-0.14

POSITIVE LOGITS

sonian

0.45

wick

0.40

ers

0.31

ies

0.30

son

0.30

ere

0.24

SON

0.21

ells

0.21

urst

0.21

aller

0.20

Activations Density 0.011%

references to people, specifically the name "Smith" and related proper nouns

No Comments

No Known Activations