INDEX

Explanations

references to specific locations, particularly states and fictional towns

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

ëĦĲ

-0.17

uls

-0.16

 Calgary

-0.15

 Alberta

-0.15

 Edmonton

-0.15

 Saskatchewan

-0.15

actus

-0.15

Ð½ÐµÐ½

-0.14

ipzig

-0.14

POSITIVE LOGITS

 Rhode

0.57

RI

0.47

 Providence

0.43

 Narr

0.42

Rh

0.40

RI

0.38

Narr

0.36

 Warwick

0.36

 Newport

0.36

RIPT

0.36

Activations Density 0.029%

references to specific locations, particularly states and fictional towns

No Comments

No Known Activations