INDEX

Explanations

references to athletic achievements and team honors

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

isson

-0.17

ica

-0.15

å¤ªéĥİ

-0.15

 Jaune

-0.15

aket

-0.14

beh

-0.14

enstein

-0.14

ihar

-0.14

abad

-0.13

ogan

-0.13

POSITIVE LOGITS

Nx

0.16

Hub

0.14

 underlying

0.14

 yÃ¶

0.14

 PROGMEM

0.13

iaz

0.13

 Ukra

0.13

å¢ĵ

0.13

 Progressive

0.13

 undone

0.13

Activations Density 0.007%

references to athletic achievements and team honors

No Comments

No Known Activations