INDEX

Explanations

specific character sequences or symbols, possibly indicating formatting or encoding elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 amnesty

-0.15

afari

-0.14

emos

-0.13

Ð¾Ð»Ð°

-0.13

Abort

-0.13

pga

-0.13

 Ð·Ð³Ð¾Ð´

-0.13

ovid

-0.13

 storage

-0.13

 Titans

-0.13

POSITIVE LOGITS

arm

0.21

 comando

0.20

dif

0.18

 campo

0.18

 blind

0.18

fan

0.18

acco

0.18

 patt

0.18

sch

0.17

Fan

0.17

Activations Density 0.004%

specific character sequences or symbols, possibly indicating formatting or encoding elements

No Comments

No Known Activations