INDEX

Explanations

references to proportions, particularly related to half or the majority of a group or population

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

amak

-0.16

amma

-0.15

reh

-0.14

imas

-0.14

iman

-0.14

 Hart

-0.14

simp

-0.13

anten

-0.13

eler

-0.13

rew

-0.13

POSITIVE LOGITS

ptom

0.14

 Noir

0.14

 Junk

0.14

utas

0.13

-inverse

0.13

 Smoke

0.13

phis

0.13

INES

0.13

aque

0.13

Activations Density 0.024%

references to proportions, particularly related to half or the majority of a group or population

No Comments

No Known Activations