INDEX

Explanations

components related to formal presentations or announcements

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

reno

-0.15

OffsetTable

-0.15

onas

-0.15

ylko

-0.14

jvu

-0.14

Å¾iv

-0.14

ceans

-0.14

 Ð²Ðº

-0.13

_sampler

-0.13

eson

-0.13

POSITIVE LOGITS

 activities

0.21

 certain

0.16

 ideas

0.16

 items

0.16

 activity

0.16

 ways

0.15

 similar

0.15

 performances

0.15

 possible

0.15

 different

0.15

Activations Density 0.147%

components related to formal presentations or announcements

No Comments

No Known Activations