INDEX

Explanations

words and phrases related to purpose and functionality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Bron

-0.17

 Blink

-0.16

iez

-0.15

ambre

-0.15

 Bentley

-0.15

 Bend

-0.15

irit

-0.15

Bounding

-0.14

 Bloom

-0.14

 Bowman

-0.14

POSITIVE LOGITS

 build

0.75

 building

0.74

build

0.70

.build

0.67

 builds

0.66

-build

0.65

-building

0.64

building

0.63

 BUILD

0.63

Build

0.63

Activations Density 0.343%

words and phrases related to purpose and functionality

No Comments

No Known Activations