INDEX

Explanations

words indicating additional or supporting information

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

atz

-0.15

bands

-0.15

ãĤŃãĥ³ãĤ°

-0.14

ÑĨÐ²ÐµÑĤ

-0.14

licht

-0.14

Serial

-0.14

atem

-0.14

ruk

-0.14

-0.13

ripe

-0.13

POSITIVE LOGITS

andler

0.20

-INF

0.19

 Feld

0.17

Ð¹Ð½

0.16

ison

0.16

iang

0.15

ertoire

0.15

 baÄŁlantÄ±lar

0.15

ideo

0.15

itre

0.14

Activations Density 0.137%

words indicating additional or supporting information

No Comments

No Known Activations