INDEX

Explanations

phrases indicating contradiction or critique in arguments

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

AtPath

-0.16

Ì§

-0.15

also

-0.14

½

-0.14

 ÑĤÐ°ÐºÐ¶Ðµ

-0.14

eger

-0.14

 also

-0.14

-0.13

 anche

-0.13

 surely

-0.13

POSITIVE LOGITS

 basically

0.26

Basically

0.24

 essentially

0.24

 Basically

0.23

 Essentially

0.20

Ess

0.18

å°±æĺ¯

0.17

tl

0.17

 kÄ±s

0.17

lesi

0.17

Activations Density 0.133%

phrases indicating contradiction or critique in arguments

No Comments

No Known Activations