INDEX

Explanations

references to moral conflicts and behavioral warnings

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

ected

-0.17

oret

-0.15

aster

-0.15

erce

-0.15

atin

-0.14

gr

-0.14

rait

-0.14

etto

-0.14

anto

-0.14

once

-0.14

POSITIVE LOGITS

_nl

0.15

 Hear

0.15

ORIZONTAL

0.15

abd

0.15

PEND

0.14

ipur

0.14

.assertIs

0.14

orb

0.14

tid

0.14

OV

0.14

Activations Density 0.044%

references to moral conflicts and behavioral warnings

No Comments

No Known Activations