INDEX

Explanations

possessive or relational phrases often associated with authority or ownership

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

agt

-0.19

ucks

-0.16

 Material

-0.15

warm

-0.15

 update

-0.15

Kam

-0.14

arias

-0.14

æŁĶ

-0.14

-0.13

avana

-0.13

POSITIVE LOGITS

eson

0.16

 cryst

0.16

tron

0.16

bac

0.16

_DEFINE

0.15

bor

0.14

setQuery

0.14

ÙĤØ¯

0.14

amen

0.14

icontrol

0.14

Activations Density 0.007%

possessive or relational phrases often associated with authority or ownership

No Comments

No Known Activations