INDEX

Explanations

references to hierarchical levels or classifications

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

efe

-0.18

oggled

-0.16

sse

-0.15

sj

-0.15

srv

-0.14

ismic

-0.14

sen

-0.14

eph

-0.14

enticate

-0.14

sko

-0.14

POSITIVE LOGITS

most

0.43

ech

0.29

 reaches

0.27

cased

0.27

archy

0.27

MOST

0.27

-middle

0.26

class

0.25

-tier

0.25

case

0.23

Activations Density 0.034%

references to hierarchical levels or classifications

No Comments

No Known Activations