INDEX

Explanations

references to nationality or citizenship, particularly American and English identities

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

inspace

-0.16

zburg

-0.16

bate

-0.15

oyer

-0.14

vinc

-0.14

WithType

-0.14

Ade

-0.14

cord

-0.14

atk

-0.14

cdecl

-0.14

POSITIVE LOGITS

avern

0.18

alth

0.15

ADDE

0.15

ukan

0.14

bench

0.14

ÎºÏĮ

0.13

ran

0.13

kÃ¡

0.13

aven

0.13

appen

0.13

Activations Density 0.015%

references to nationality or citizenship, particularly American and English identities

No Comments

No Known Activations