INDEX

Explanations

references to crime or criminal activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

-Cs

-0.15

repid

-0.14

tura

-0.14

xbb

-0.13

 Mess

-0.13

vailability

-0.13

 Dani

-0.13

EINVAL

-0.13

ebi

-0.13

ioni

-0.13

POSITIVE LOGITS

rogram

0.14

itespace

0.14

itm

0.14

izzo

0.14

agrams

0.13

om

0.13

linear

0.13

Xem

0.13

grams

0.13

 quÃ¢n

0.13

Activations Density 0.027%

references to crime or criminal activity

No Comments

No Known Activations