INDEX

Explanations

references to reading and the experience of engaging with written content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ì¶ľ

-0.16

Wor

-0.15

iske

-0.15

resher

-0.14

quin

-0.14

ubes

-0.14

nesc

-0.13

 Template

-0.13

newline

-0.13

qv

-0.13

POSITIVE LOGITS

 reading

0.41

 read

0.37

 reads

0.33

reading

0.32

éĺħè¯»

0.32

Reading

0.31

 Reading

0.30

è¯»

0.30

read

0.29

-read

0.28

Activations Density 0.238%

references to reading and the experience of engaging with written content

No Comments

No Known Activations