INDEX

Explanations

references to line numbers and operations in code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

inker

-0.17

ingleton

-0.15

QueryBuilder

-0.15

è¦Ĩ

-0.14

ept

-0.13

ç¸

-0.13

nev

-0.13

ãģ¡ãĤĥãĤĵ

-0.13

yw

-0.13

 Till

-0.13

POSITIVE LOGITS

eno

0.17

onds

0.16

ofi

0.15

ØŃÛĮ

0.15

ime

0.14

gar

0.14

dtype

0.14

datatype

0.14

Activations Density 0.039%

references to line numbers and operations in code

No Comments

No Known Activations