INDEX

Explanations

patterns of statistical data or numerical results

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

á»§

-0.16

CLUDED

-0.15

ube

-0.15

ardon

-0.14

ushi

-0.14

jen

-0.14

cola

-0.14

ÌĨ

-0.14

atorium

-0.14

illet

-0.14

POSITIVE LOGITS

oa

0.15

iken

0.14

 Ø§ÙĦØ±

0.13

sod

0.13

 tops

0.13

 respect

0.13

UBL

0.13

 tens

0.13

 Crystal

0.12

Activations Density 0.014%

patterns of statistical data or numerical results

No Comments

No Known Activations