INDEX

Explanations

references to a specific individual named Hart

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

hoot

-0.19

routine

-0.17

itaire

-0.16

rael

-0.15

UREMENT

-0.15

rine

-0.15

andering

-0.15

ront

-0.14

rons

-0.14

bones

-0.14

POSITIVE LOGITS

nett

0.23

igan

0.23

kop

0.19

nell

0.19

sock

0.18

mann

0.18

mut

0.17

well

0.17

wig

0.17

kop

0.17

Activations Density 0.004%

references to a specific individual named Hart

No Comments

No Known Activations