INDEX

Explanations

terms related to death, injury, and damage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

enne

-0.20

emann

-0.15

ise

-0.15

eller

-0.15

sockets

-0.15

eria

-0.14

zv

-0.14

ike

-0.14

 Leban

-0.14

enment

-0.14

POSITIVE LOGITS

/null

0.17

McB

0.17

apper

0.17

ulus

0.15

&e

0.15

acon

0.14

/exp

0.14

-from

0.14

simp

0.14

Ã¥l

0.14

Activations Density 0.134%

terms related to death, injury, and damage

No Comments

No Known Activations