INDEX

Explanations

phrases that indicate personal experiences or long-term familiarity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

loor

-0.16

ombat

-0.16

gne

-0.15

.dependencies

-0.15

dream

-0.14

steen

-0.14

ĭè¯ķ

-0.14

ITA

-0.14

ampion

-0.14

 Trilogy

-0.14

POSITIVE LOGITS

 myself

0.17

 professionally

0.16

 privileged

0.16

 ÐĹÐ°Ð¼

0.16

ippo

0.15

ble

0.15

 worse

0.15

 both

0.15

rish

0.14

 priv

0.14

Activations Density 0.151%

phrases that indicate personal experiences or long-term familiarity

No Comments

No Known Activations