INDEX

Explanations

terms and phrases related to searching for information and data management

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

figcaption

-0.14

leur

-0.13

rof

-0.13

ÏĨÎ¿ÏģÎ¬

-0.13

.WriteAll

-0.12

áº¡

-0.12

ControlEvents

-0.12

oday

-0.12

 sigmoid

-0.12

defgroup

-0.12

POSITIVE LOGITS

 search

1.20

 Search

1.05

search

1.05

Search

1.01

-search

0.99

 searches

0.96

_search

0.95

.search

0.94

 SEARCH

0.92

æĲľç´¢

0.90

Activations Density 0.378%

terms and phrases related to searching for information and data management

No Comments

No Known Activations