INDEX

Explanations

themes related to deception or the appearance of things being misleading

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

iero

-0.15

 ring

-0.15

eros

-0.14

reet

-0.14

roz

-0.14

isma

-0.14

aland

-0.14

imal

-0.14

phia

-0.14

ÑĦÐ¾ÑĢ

-0.13

POSITIVE LOGITS

oj

0.15

impse

0.15

orsch

0.14

ey

0.14

-windows

0.13

emente

0.13

auf

0.13

uther

0.13

Station

0.13

eb

0.13

Activations Density 0.095%

themes related to deception or the appearance of things being misleading

No Comments

No Known Activations