INDEX

Explanations

the word "right" in various contexts of selection or correctness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

ESH

-0.16

iggs

-0.15

rz

-0.14

 Bang

-0.14

ogne

-0.14

llib

-0.13

lyph

-0.13

 Carry

-0.13

ards

-0.13

 carry

-0.13

POSITIVE LOGITS

arch

0.15

ellan

0.15

orch

0.15

roi

0.14

 kind

0.14

 Stuff

0.14

owitz

0.14

 klik

0.13

 spons

0.13

agan

0.13

Activations Density 0.026%

the word "right" in various contexts of selection or correctness

No Comments

No Known Activations