INDEX

Explanations

expressions of comparison and contrasting opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

esign

-0.15

igne

-0.14

odable

-0.14

ãģĭãĤīãģ®

-0.14

izedName

-0.14

erland

-0.14

_UNUSED

-0.13

ç½

-0.13

ÏģÎŃ

-0.13

olocation

-0.13

POSITIVE LOGITS

acas

0.20

nip

0.16

endo

0.14

imo

0.14

pro

0.14

anners

0.14

anzi

0.14

PJ

0.14

anos

0.13

Activations Density 0.055%

expressions of comparison and contrasting opinions

No Comments

No Known Activations