INDEX

Explanations

programming constructs and method calls

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

ForResult

-0.17

oud

-0.17

@brief

-0.15

odu

-0.15

viar

-0.14

 nullptr

-0.14

ÙĪØ¯ÛĮ

-0.14

 Latter

-0.14

pto

-0.14

tout

-0.14

POSITIVE LOGITS

int

0.35

 final

0.22

 String

0.21

 boolean

0.20

 switch

0.19

int

0.18

int

0.18

final

0.17

itty

0.17

try

0.17

Activations Density 0.059%

programming constructs and method calls

No Comments

No Known Activations