INDEX

Explanations

variations of the word "ideology."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ond

-0.19

inand

-0.16

ÑıÐ¶

-0.16

iger

-0.15

etro

-0.15

OND

-0.15

train

-0.15

resse

-0.15

rou

-0.15

xon

-0.15

POSITIVE LOGITS

als

0.26

pend

0.23

ologically

0.20

ally

0.20

ALLY

0.19

yal

0.18

yll

0.18

ntity

0.18

ologies

0.18

ALS

0.18

Activations Density 0.007%

variations of the word "ideology."

No Comments

No Known Activations