INDEX

Explanations

code-related constructs and types

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ãģıãĤĮ

-0.16

rez

-0.15

Kir

-0.15

 Hercules

-0.15

Persona

-0.14

wed

-0.14

urette

-0.14

 Milo

-0.14

eder

-0.14

iren

-0.14

POSITIVE LOGITS

 class

0.17

ï»¿using

0.17

stands

0.16

 definitions

0.16

class

0.15

LEM

0.15

_defs

0.15

errupted

0.15

 definition

0.14

IOUS

0.14

Activations Density 0.148%

code-related constructs and types

No Comments

No Known Activations