INDEX

Explanations

phrases that indicate advanced or innovative techniques or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ako

-0.17

ileo

-0.16

oun

-0.14

sted

-0.14

EAR

-0.14

rico

-0.14

 impost

-0.14

oves

-0.14

pies

-0.14

eration

-0.13

POSITIVE LOGITS

 edge

0.21

edge

0.21

-edge

0.20

Age

0.17

age

0.16

RetVal

0.15

_edge

0.15

erg

0.15

gers

0.15

ãĤ¨

0.14

Activations Density 0.005%

phrases that indicate advanced or innovative techniques or concepts

No Comments

No Known Activations