INDEX

Explanations

terms related to fragments and snapshots in data structures

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ÌĨ

-0.17

boys

-0.17

oted

-0.17

boy

-0.17

yo

-0.16

fully

-0.16

ãĥ³ãĥĳ

-0.16

upo

-0.15

umer

-0.15

sed

-0.14

POSITIVE LOGITS

al

0.17

xito

0.17

ments

0.17

 pieces

0.15

à¹ģà¸£à¸ģ

0.15

stakes

0.15

TS

0.15

 halinde

0.15

visa

0.15

alist

0.15

Activations Density 0.146%

terms related to fragments and snapshots in data structures

No Known Activations