INDEX

Explanations

expressions of bias and subjective opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

èĮ

-0.16

 Appropri

-0.16

igu

-0.15

Hir

-0.14

 rast

-0.14

(Void

-0.14

Ð¾Ð½ÑĤ

-0.14

ermen

-0.14

æīķ

-0.13

engan

-0.13

POSITIVE LOGITS

 biased

0.28

 bias

0.24

 Bias

0.20

biased

0.19

 biases

0.18

bias

0.18

ince

0.18

 partisan

0.18

ffi

0.16

.toObject

0.16

Activations Density 0.186%

expressions of bias and subjective opinions

No Comments

No Known Activations