INDEX

Explanations

references to programming code and coding-related terminology

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

Codes

-0.18

ly

-0.18

ness

-0.18

_codes

-0.18

la

-0.17

most

-0.17

Ð»Ð°

-0.16

Code

-0.16

ãĥ³ãĥĸ

-0.16

POSITIVE LOGITS

base

0.34

æ®µ

0.29

bases

0.26

block

0.25

pend

0.25

pen

0.24

-sn

0.24

blocks

0.24

_sn

0.23

 snippet

0.23

Activations Density 0.026%

references to programming code and coding-related terminology

No Comments

No Known Activations