INDEX

Explanations

phrases indicating prior experience or background

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

.joda

-0.15

 CONST

-0.14

sembl

-0.13

sembly

-0.13

IPH

-0.13

ony

-0.13

isp

-0.13

argon

-0.12

rastructure

-0.12

istribution

-0.12

POSITIVE LOGITS

 moving

0.20

 joining

0.19

 leaving

0.19

 immigr

0.18

 Join

0.18

moving

0.17

 assuming

0.17

 arriving

0.17

 coming

0.17

 paren

0.16

Activations Density 0.039%

phrases indicating prior experience or background

No Comments

No Known Activations