INDEX

Explanations

phrases indicating registration or subscription for free access

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

reten

-0.17

cono

-0.15

ayah

-0.15

ovou

-0.14

ë¶

-0.14

rott

-0.14

antha

-0.14

Ã³b

-0.14

 crossorigin

-0.14

orgot

-0.14

POSITIVE LOGITS

 partial

0.22

 partially

0.21

partial

0.21

 Partial

0.19

 partly

0.19

Partial

0.17

ghan

0.17

ee

0.16

.partial

0.15

.lex

0.15

Activations Density 0.002%

phrases indicating registration or subscription for free access

No Comments

No Known Activations

phrases indicating registration or subscription for free access

No Comments

No Known Activations