INDEX

Explanations

elements of coding or programming 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

Not in Any Lists

No Comments

Negative Logits

ÙĪØ²

-0.15

amd

-0.15

nement

-0.15

umba

-0.15

enburg

-0.14

ypical

-0.13

reducers

-0.13

usch

-0.13

isted

-0.13

ish

-0.13

POSITIVE LOGITS

arta

0.16

aviolet

0.15

 Horizon

0.15

 !***

0.14

iolet

0.14

ÑģÐ»Ð¾Ð²

0.14

æĹıèĩªæ²»

0.14

argo

0.14

wake

0.14

EncodingException

0.14

Activations Density 0.048%

elements of coding or programming structures

No Comments

No Known Activations