INDEX

Explanations

statements that emphasize contradictions or questions around societal norms and roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

achu

-0.15

ReadStream

-0.15

Occurs

-0.14

hti

-0.14

antt

-0.14

iaux

-0.14

bose

-0.13

_SUFFIX

-0.13

miner

-0.13

 DeepCopy

-0.13

POSITIVE LOGITS

#ad

0.20

undy

0.15

 Shepard

0.14

illo

0.14

 Interr

0.13

ape

0.13

Of

0.13

datable

0.13

Of

0.13

Ä°T

0.13

Activations Density 0.247%

statements that emphasize contradictions or questions around societal norms and roles

No Comments

No Known Activations