INDEX

Explanations

references to dates and locations in a news context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

batim

-0.15

quila

-0.15

rien

-0.15

urette

-0.14

iese

-0.14

 Jobs

-0.14

ynth

-0.14

ostel

-0.14

 jobs

-0.14

erk

-0.14

POSITIVE LOGITS

(PR

0.16

ĵĺ

0.16

APH

0.15

andelier

0.15

ezier

0.14

VAR

0.14

aho

0.14

 ê°Ī

0.14

å¨

0.14

 Wich

0.14

Activations Density 0.003%

references to dates and locations in a news context

No Comments

No Known Activations