INDEX

Explanations

instances of negation and conditional statements regarding beliefs or assumptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

nock

-0.16

ãĥ³ãĥĦ

-0.16

iaux

-0.16

meni

-0.16

nze

-0.15

prene

-0.15

ityEngine

-0.15

imax

-0.15

nox

-0.15

gamber

-0.15

POSITIVE LOGITS

yl

0.16

alg

0.15

els

0.15

0.14

lun

0.14

iddy

0.14

ledge

0.14

 Congress

0.14

ylim

0.13

Activations Density 0.012%

instances of negation and conditional statements regarding beliefs or assumptions

No Comments

No Known Activations