INDEX

Explanations

references to the concept of truth

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

list

-0.15

utton

-0.15

 Reign

-0.14

dar

-0.14

oot

-0.14

aggio

-0.13

ifton

-0.13

iao

-0.13

ilon

-0.13

oo

-0.13

POSITIVE LOGITS

fully

0.24

fulness

0.21

ãģªãĤĭ

0.17

soever

0.17

/false

0.17

ifier

0.16

.djangoproject

0.16

ushima

0.16

flen

0.16

LayoutPanel

0.15

Activations Density 0.020%

references to the concept of truth

No Comments

No Known Activations