INDEX

Explanations

actions related to selection or choice

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

Ð²Ð¸Ð¶

-0.16

avig

-0.16

ion

-0.16

tracker

-0.15

od

-0.14

chez

-0.14

enia

-0.14

ãĤ¤ãĥ³ãĥĪ

-0.14

Ð´Ð°ÑĤÑĮ

-0.14

æĪĲç«ĭ

-0.14

POSITIVE LOGITS

.pick

0.31

 picking

0.28

 picked

0.26

 PICK

0.25

 pick

0.25

 picks

0.25

pick

0.24

 lá»įc

0.24

 apart

0.23

picked

0.22

Activations Density 0.017%

actions related to selection or choice

No Comments

No Known Activations